CEE_20_04

Control, Instrumentation and Automation in the Process and Manufacturing Industries April 2020

www.controlengeurope.com

Two disciplines – one device

Having a vision for AI and deep learning

Hope for the best but always prepare for the worst

5G benefits and barriers

HORNER OCS HMI, PLC, Network and I/O THE “ALL IN ONE” CONTROLLER Range of Display Options Available 2.2" to 22"

Micro SD (32GB)

10/100 Ethernet

USB 2.0 Host USB 2.0 on-the-go

Digital & Analogue Inputs

Digital & Analogue Outputs

RS-232 & RS-485 Serial Ports CAN Port High Performance Logic Controller

SEE OUR COMPLETE RANGE ONLINE AT WWW. HORNERAUTOMATION.EU/OCSRANGE

• Inbuilt I/O Suits Many Applications • Single IEC61131 Programming Software for Visualisation and Logic • Powerful WebServer

• Battery Operation (9-30V) • Extensive Networking Capability • Customisable Options

“Designed, Developed, Assembled and Supported in Europe since 1997.”

Cork, Ireland Tel: +353 21 4321 266 Email: info@hornerautomation.eu www.hornerautomation.eu/OCSRange

CONTENTS Moving into unprecedented times

Editor Suzanne Gill suzanne.gill@imlgroup.co.uk Sales Manager Adam Yates adam.yates@imlgroup.co.uk Group Publisher Iain McLean iain.mclean@imlgroup.co.uk Production Holly Reed holly.reed@imlgroup.co.uk Dan Jago Group Publisher David May Production Manager G and C Media Studio Design

Producing this issue of Control Engineering Europe has been interesting to say the least. Over the course of the last few weeks, exhibitions across Europe have, understandably, either postponed or cancelled due to the threat posed by coronavirus.

That has made me think about what the loss of these events means to industry. Exhibitions provide a unique opportunity to see products and solutions in action, to touch and feel them, to interact with suppliers and to network with peers to find out more about the latest trends and technologies. This year, it looks like we may have to forego this opportunity and I am interested to better understand what that means for those of you who were planning to attend shows in the coming weeks and months with planned installations in mind? Will you make your purchasing decisions regardless – gathering the information you need by other means – or will your plans also be delayed? Please do let me know how these changes affect your purchasing plans and what Control Engineering Europe might be able to do to help keep you informed.

Suzanne Gill Editor – Control Engineering Europe suzanne.gill@imlgroup.co.uk

INDUSTRY REPORT

PREDICTIVE MAINTENANCE

4 Standard industrial interface for SPE set published

26 Turning the predictive maintenance dream into reality

EDITOR’S CHOICE

INDUSTRIAL ETHERNET

6 Bringing versatility to edge computing; Smart actuators for process applications

28 Big data analytics moves right to the edge with the introduction of the IEEE 802.3cg standard for industrial automation

CYBERSECURITY 12 Myth busting: best practice advice to help enhance industrial network security

EDGE COMPUTING

14 When it comes to cybersecurity we can hope for the best but must always prepare for the worst

30 What is the difference between cloud and edge computing and how can they work together for the successful implementation of IIoT solutions?

17 David Emm, principal security researcher at Kaspersky, answers questions about industrial cyber threats and vulnerabilities.

MACHINE VISION 18 Advice from industry experts about how end users can ensure that they are able to implement successful AI-based machine vision applications

WIRELESS TECHNOLOGY: 5G 22 Find out more about the role 5G is set to play in enabling digitalisation of enterprises as well as the current barriers to its deployment Control Engineering Europe is a controlled circulation journal published eight times per year by IML Group plc under license from CFE Media LLC. Copyright in the contents of Control Engineering Europe is the property of the publisher.

ISSN 1741-4237 IML Group plc Blair House, High Street, Tonbridge, Kent TN9 1BQ UK Tel: +44 (0) 1732 359990 Fax: +44 (0) 1732 770049

Control Engineering Europe

PG 8

Control Engineering (USA) Mark Hoske, Circulation Tel: +44 (0)1732 359990 Email: subscription@imlgroup.co.uk Completed print or on line registration forms will be considered for free supply of printed issues, web site access and on line services.

www.controlengeurope.com

Qualified applicants in Europe must complete the registration form at http://imlrenewals.managemyaccountonline.net to receive Control Engineering Europe free of charge.

Paid subscriptions for non-qualifying applicants are available for £113 (U.K.), £145 (Europe), £204 (rest of world); single copies £19.

April 2020

3

INDUSTRY REPORTS

Top robotic trends for 2020 From 2020 to 2022 almost two million new units of industrial robots are expected to be installed in factories around the world. The International Federation of Robotics (IFR) explains why.

“Smart robotics and automation are vital to deal with new consumer trends, demand for product variety or challenges from trade barriers”, said Dr. Susanne Bieller, general secretary of the IFR. “New technological solutions pave the way for more flexibility in production.”

Simplification, collaboration and digitalisation are believed to be the key drivers that will benefit robot implementation.

Robots get smarter: Programming and installation of robots is becoming much easier. In practice this sees digital sensors combined with smart software to allow direct teaching methods. The task that the robot arm is to perform is first executed by a human who just has to guide the robot arm through the required movements. This data is then

transformed by the software into the digital program of the robot arm. In the future, machine learning tools will further enable robots to learn by trialand-error or by video demonstration and will self-optimise their movements.

Will robots collaborate with workers: Human-robot collaboration is another important trend in robotics. Working alongside humans, modern robotic systems are able to adapt to a rapidly changing environment. The range of collaborative applications offered by robot manufacturers continues to expand. Currently, shared workspace applications are most common with the robot and worker operating alongside each other, completing tasks sequentially. Applications in which the human and the robot work at the same time on the same part are more challenging. Research and Development (R&D) focuses on methods to enable robots to respond in real-time. Just like two human workers would collaborate, the R&D teams wants the robot to adjust its motion to its environment,

allowing for a true responsive collaboration. Solutions to enable this include voice, gesture and recognition of intent from human motion.

Robots go digital: Industrial robots are the central components of digital and networked production solutions.

This makes it important for them to be able to communicate with each other – regardless of their manufacturer. The OPC Robotics Companion Specification, which has been developed by a joint working group of the VDMA and the OPC Foundation, defines a standardised generic interface for industrial robots and enables industrial robots to connect into the Industrial Internet of Things (IIoT). The digital connectivity of robots with cloud technology is also an enabler for new business models such as robot leasing.

Robots-as-a-Service has advantages that might be attractive for small and medium-sized enterprises (SMEs) – no committed capital, fixed costs, automatic upgrades and no need for highly-qualified robot operators.

Huge growth expected for 5G in manufacturing The market for 5G cellular connections in manufacturing is expected to reach US$10.8 billion by 2030, at a Compound Annual Growth Rate (CAGR) of 187%, according to global tech market advisory firm, ABI Research. It says that 5G and edge computing ‘constitute a technological leap that heralds a significant transformation of business models for all industries, including manufacturing and associated Industry 4.0 verticals’.

“To capture this value, ecosystem stakeholders will first need to evaluate how to measure the impact of 5G and edge deployments,” said Don Alusha, senior analyst at ABI Research. The current Industry 4.0 digitalisation discourse centers around conventional financial metrics

4

April 2020

such as return on investment (ROI) net profit, and cash flow as the yardstick to measure 5G and edge computing effectiveness. However, these metrics gauge profit and do not lend themselves to the factory floor.

“Therefore, Industry 4.0 ecosystem entities must consider an alternative set of measurements that look at how 5G and edge deployments aid manufacturing establish operational rules to run a plant.

They are throughput, inventory and operational expense for the incoming flow of capital, for capital located inside, and for capital going out, respectively,” explained Alusha.

These three measurements enable Industry 4.0 partners (such as ABB, Bosch and Siemens) to institute a direct connection between the 5G’s utility

www.controlengeurope.com

and what takes place on the factory floor. In turn, they will be able to use that connection to find a logical relationship between daily plant operations and the overall company’s performance. Only then, will Industry 4.0 verticals have a basis for knowing the real benefit of 5G and edge computing. “Furthermore, and equally important, is the ability to measure risk when looking to adopt 5G and edge technology assets. Discussions on new technology adoption have always been based on an assessment of risk and reward. If the reward is truly compelling, adopters will take the risk.

5G and edge offer unprecedented commercial opportunities, but they inherently constitute new technologies and therefore there is a risk attached,” said Alusha.

Control Engineering Europe

INDUSTRY REPORTS

Standard industrial interface for SPE set published The international standard for Single Pair Ethernet (SPE) interfaces in industrial applications has been published with IEC 63171-6.

IEC 63171-6 (Industrial Style) is a complete standard document with all necessary specifications and test sequences, which is incorporated into current SPE cabling standards of the ISO/IEC 11801-x series of standards for structured cabling.

Single Pair Ethernet (SPE) describes the transmission of Ethernet over only one pair of copper wires. In addition to data transmission via Ethernet, SPE also enables a simultaneous power supply of terminal devices via PoDL – Power over Data Line. Previously, this required two pairs for Fast Ethernet (100MB) and four pairs for Gigabit Ethernet.

The implementation of SPE in the ISO/IEC 11801 documents is important because only in this standard are the cabling channels described with all necessary parameters – length, number of connections, bandwidth and the complete set of transmission parameters including NEXT, FEXT, shielding

properties etc – with relation to the environment and can therefore be metrologically verified after installation.

This connection of component standards to connectors and cables should provide users of SPE with clear guidelines for the construction and testing of appropriate transmission links.

Turn Data Into Actionable Insights Reduce complexity, improve quality and enhance energy management with AVENTICS™ IIoT solutions from Emerson.

From measuring critical parameters to detecting leaks to improving predictive maintenance, we make data visible for faster, better-informed decision-making.

Rely on Emerson for intelligent pneumatics that add value to your business.

Learn more at: Emerson.com/AVENTICS

The Emerson logo is a trademark and a service mark of Emerson Electric Co. © 2020 Emerson Electric Co.

Emerson_Master_IIoT_20_178x121_EN.indd 1

3/11/2020 8:31:47 AM

EDITOR’S CHOICE

Bringing versatility to edge computing Red Lion Controls has introduced the FlexEdge Intelligent Edge Automation Platform which is said to bring versatility to edge computing, while offering productivity and efficiency gains from digital transformation initiatives thanks to its simplicity.

The FlexEdge platform carries several certifications that make it suitable for use in oil and gas, water and wastewater, maritime, hazardous locations, and factory automation applications.

The modular architecture of this communication gateway boasts a variety of wireless and wired communication options that allow for connection with any industrial communication requirement, regardless of protocol or manufacturer.

FlexEdge offers a form factor

and platform that allows for rapid adaptation as application needs change.

With FlexEdge it is said to be possible to connect new and existing devices, reducing overall downtime and

Easy migration without detours

Siemens is expanding its distributed I/O system with new modules for the Simatic ET 200MP/Simatic S7-1500 IOs as well as a backplane bus. A total of four new multi-channel digital modules, with 64 channels each and an installation width of 35mm, enables users to achieve spacesaving and cost-efficient implementation of a high number of channels in the control cabinet. Particularly existing systems on the basis of Simatic S7-300/ ET 200M are easy to migrate to the new Simatic S7-1500 systems with the new portfolio. Hot-swapping of modules is also possible in combination with the new active backplane bus for the

6

April 2020

Simatic S7-1500 or Simatic ET 200MP. This means that users can replace any faulty Simatic ET 200MP modules during the CPU runtime, i.e. during ongoing plant operation, while non-affected modules remain operational, thus increasing plant availability. The new portfolio makes the Simatic ET 200MP suitable for use in the process environment together with Simatic S7-1500R/H systems since it promotes high channel density and high levels of availability.

With the Simatic TOP Connect system cabling, users also benefit from short installation times for the new digital modules since no tools are required.

www.controlengeurope.com

allowing for frictionless, future-proof scalability. Its flexibility allows the selection of one platform for any edge requirement.

For automation engineers, it offers a single platform that can solve most challenges while decreasing the number of deployed devices to reduce application complexity. For networking engineers, it offers a robust, industrialgrade networking gateway deployable as a wired or cellular solution to securely connect different networks.

With multiple isolated serial ports, Ethernet, optional Wi-Fi and cellular communication sleds, optional fieldinstallable I/O, and a certified, robust enclosure, the FlexEdge solution is said to ensure optimal operation in almost any industrial environment with demanding requirements.

Smart actuators for process applications The new PROFOX electric valve actuator is said to be suited for process industry applications which require fast and precise positioning, flexibility and future-proof interfaces.

Built-in intelligence makes the actuator suited to both simple open-close duty and challenging modulating applications. Motor speed is adjustable, ensuring fast and precise positioning without overrun, while soft start and soft stop functionality increases valve lifetime. Diagnostics capabilities help ensure long-term availability and future-proof these IP67-rated actuators for IIoT applications.

The PROFOX will work with gate, butterfly, ball and globe valves.

Currently, the range includes multi-turn versions for torques of 10 – 100 Nm and part-turn versions delivering 32 – 600 Nm. Control Engineering Europe

AI 05-19.000.L1

C_ber Securi_y Incomplete security?

Our 360° cybersecurity solutions protect your industrial network!

Industrial control and automation systems are increasingly exposed to cyber risks and unintentional changes due to the growing networking of these systems and their connection to the Internet. By means of secure products, services, and industrial solutions, Phoenix Contact helps you protect your systems and safeguard your know-how. Don’t hesitate to contact us for more information and advice.

For additional information call +49 5235 3-00 or visit phoenixcontact.com

COVER STORY

Two disciplines – one device The integration of a fail-safe motion control system, multi-axis drive control and powerful communication interfaces and technology I/Os in one device saves valuable control cabinet space and facilitates the implementation of particularly challenging machine concepts extremely efficiently – including modularisation and optimum scalability.

T

he requirements for automation and drive technology in today’s production machines are demanding: In addition to the ‘classic’ control technology, they are predominantly equipped with a multitude of highly dynamic servo drives, which must be operated energyefficiently in perfect synchronisation with one another, by way of cam and gear synchronous operations, for example.

Complementing the typical functions of a PLC (e.g. logic processing), elaborate motion control functions are crucial – such as for positioning, synchronism and cam discs, through to path interpolation with transformations for handlings kinematics.

Technological I/Os are also required for position-related acquisition and issuing of binary signals.

Flexible scalable solutions, which permit the simple implementation even of modular machine concepts, are essential for rapid reactions to changes in the market. Moreover, this includes integrated safety concepts as protection for the machine and personnel, powerful communication interfaces for the incorporation of drive, I/O and HMI systems, as well as integration in system networks, MES systems, and cloud services.

In addition to classic PLCs and PCbased automation solutions, drive-based control platforms in particular have also established themselves for production machines which are reliant on motion control. The automation and drive

8

April 2020

drives to be operated directly at the controller. Should additional drives be required, the drive quantity structure can be expanded by way of Sinamics S120 Control Units (CU320-2) or Sinamics S210 single-axis AC/AC converter via Profinet, for example. Standardised data exchange via Profidrive also allows the integration of drives from other manufacturers.

The Drive Controller is available in multiple versions for axis quantity structures of more than 100 axes.

With the ‘cross-PLC synchronous operation’ function integrated in all S7-1500 technology CPUs, cam and gear synchronous operations are also possible across all CPUs, ensuring that the axis quantity structures are practically limitless. In addition to the performance distribution to multiple CPUs, the crossPLC synchronous operation also allows the uncomplicated implementation of modular automation concepts with ‘autonomous function units’.

In order to consider the safety aspects of producing machines, safe monitoring of all movements is also possible.

technology of many production machines often have to make do with cramped recesses in the machine bed – therefore, particularly compact and space-saving control cabinet solutions are required.

With the Simatic Drive Controller, Siemens AG has expanded the product portfolio of the Simatic S7-1500 Advanced Controller to include a drive-based version. The new controller optimises the integration of the Simatic S7-1500 control system and the Sinamics S120 drive system. It integrates motion Safely increase productivity control, technology, PLC and safety Intelligent safety concepts are not only functionalities directly in the modular and designed to protect personnel and highly dynamic multi-axis drive system machinery, they also form the basis for – without the need for any additional high machine throughput and minimal space for the control system in the control downtimes. cabinet.

Danger zones can be protected by The integrated interfaces and safety light grids, for example, and drives technology I/Os are available in all operated safely by applying the Safe performance classes of the controller.

Direction (SDI) function, where a danger This simplifies scaling based purely on zone can be made accessible to the performance and allows for efficient operator provided the machine moves in implementation of compact and modular automation and drive solutions. Engineering of the controller is performed via the TIA Portal with Simatic Step 7 and Sinamics Startdrive.

The integrated drive control of the Drive Controller allows up to 6 servo, Control cabinets in the machine bed. Space-saving solutions for the 6 vector or 12 V/f control and drive technology are indispensable in this regard.

www.controlengeurope.com

Control Engineering Europe

COVER STORY

For a space-saving layout, the power units of the multi-axis drive system can be connected directly without any additional Control Unit (CU) for the Drive Controller. Applications can be flexible scaled and modularised via Profinet and cross-PLC synchronous operation.

a safe direction away from the operator, ensuing that the user can safely convey material into or out of the working area, for example.

The fail-safe controller features a safety function for the control system and integrated drive control, facilitating integrated solutions which also fulfill the most demanding requirements with regard to the safety of personnel and machinery. This safety functionality can be flexibly adapted to individual requirements using the freely programmable F-CPU. The wizard-guided safety acceptance test for the Startdrive drive engineering software leads the user step-by-step through the acceptance process and checks whether the safety functions have been parameterised correctly and are properly executed in the respective application. A standardcompliant acceptance certificate can then be created automatically for documentation.

Effective communication based on Profinet with Profisafe reduces the time and effort required for cabling as compared with conventional wiring with safety relays.

Future-oriented communication State-of-the-art automation solutions necessitate future-oriented communication solutions. These requirements are challenging: Large data volumes are to be exchanged with higherControl Engineering Europe

level systems and reproducible processing of I/O signals is required on the field level with the shortest terminal-terminal times, including fail-safe communication.

Motion control applications represent a particular challenge here with regard to communication: In addition to short cycle times for maximum productivity and product quality, a fast deterministic data exchange with drive systems and I/O systems or between machine modules is absolutely essential.

Profinet with Isochronous Real-Time (IRT) is used here for the Drive Controller.

In all, there are three Profinet interfaces and one Profibus interface available, allowing the machine network to be separated from the plant network from an addressing point of view, whilst also allowing user-friendly connection of an MES system or cloud services, for example. As well as the option of distributed connection of the I/Os via the communication interfaces, special onboard I/Os are also available on the Drive Controller.

Technology I/Os for motion control Motion control applications are not only limited to the actual traversing of axes.

They also include technological functions which are closely linked with the motion control. To determine the product position on a conveyor belt, for example, measuring inputs record the axis position with maximum precision when the edge

www.controlengeurope.com

rises and/or falls at a digital input. Further, Cam outputs switch a digital output in accordance with the respective position.

These technological I/Os have a high distribution rate for product machines, however, these are often only required in lesser quantities. In order that these applications can be implemented as efficiently as possible (without additional modules), the controller features a further 8 technology I/Os in addition to 8 measuring inputs on the drive side, the respective functions of which can be parameterised channel-selectively.

To ensure maximum precision in the µs range, the inputs feature a parametrisable input filter and the high-speed outputs feature special output drivers, which within 1 µs switch with extreme edge steepness, and therefore with reproducibility. The attainable accuracy is thus essentially only determined by the sensors or actuators used.

Production machines require highperformance automation solutions with minimum spatial requirements. To satisfy these requirements, the Simatic Drive Controller combines a fail-safe technology CPU with a multi-axis drive control requiring very little space. The, integrated interfaces and technology I/Os are available uniformly in all performance classes so that scaling is particularly easy to perform and automation solutions can be implemented with extreme efficiency.

The cross-PLC synchronous operation also supports the implementation of modular machine concepts, as well as performance distribution to multiple CPUs for particularly extensive axis quantity structures.

Providing maximum functionality in the smallest of spaces, the Drive Controller is an attractive alternative for the efficient implementation of production machines with comprehensive motion control functionality.

The Drive Controller is embedded in Totally Integrated Automation and therefore benefits from the holistic digitalisation solutions from Siemens AG. www.siemens.com//drive-controller April 2020

9

PROCESS CONTROL

Overcoming your process control migration woes Lynn Njaa says that there are some important questions that you need to answer when considering migrating from a legacy distributed control system.

W

hat manufacturer doesn’t cringe at the thought of migrating legacy control system technology?

Most control systems were installed or updated in the late 80s and 90s and are overwhelmingly fragmented and fraught with challenges. Taking a waitand-see approach is no longer a viable option, as manufacturers face new standards and regulatory compliance mandates, along with disparate and obsolete systems requiring hard-to-find original equipment manufacturer (OEM) spare parts and higher support costs.

Compounding these challenges are

10

April 2020

associated safety and cybersecurity risks in operating and maintaining legacy systems.

With all the new innovative technologies on the market, now is a good time to re-evaluate processes and systems and look at opportunities to improve operational efficiency and performance to stay competitive.

For many manufacturers, tackling a project of this magnitude, with limited bandwidth and resources, is no easy feat. It’s also not easy deciding whether to replace systems piecemeal or start over with new systems. Upfront planning and a qualified team are key

www.controlengeurope.com

to moving into the execution phase of an upgrade or migration project for a distributed control system (DCS), process control system (PCS), or programmable logic controller (PLC)-based system.

Plan and execute First, before deciding on which upgrade or migration path to take, it is important to take a holistic look at the entire facility’s operations.

Regardless of a facility’s size, a solid plan from conception of the migration through startup must be considered for a successful project. A disciplined approach helps to define a scope that

Control Engineering Europe

PROCESS CONTROL aligns with business goals and objectives and to outline facility requirements.

This front-end loading (FEL) engineering effort helps ensure an execution plan and schedule are in place to keep the project(s) on time and on budget. It also supports defendable control system solutions to determine the best platform, identify and define any risks, and provide accurate and justifiable cost estimates. The result is a successful project execution plan with a functional platform in place for operational efficiency and profitability.

Engage a qualified team Where bandwidth and resource limitations are an issue do consider engaging third-party control system specific experts up front. Using them throughout the project will increase project success. Experienced guidance will pay dividends if obtained during the conceptual phase of the migration.

Waiting to engage an external partner later in the project means many aspects of the scope will already have been defined, based on assumptions made up front, that may or may not have included an all-inclusive view of the entire site’s needs. Changing these decisions late in the project will result in higher costs and schedule delays compared with getting them set correctly during the initial FEL phase.

A platform-independent partner will

deliver unbiased practical experience working on a variety of manufacturing processes and technologies. This qualified team of experts will utilise best practices and successfully help lead the project through the execution and implementation phases. It is best to work with a team early in the FEL stage to avoid delayed schedules and higher costs later.

Additionally, an ideal partner is one who is in it for the long haul and knows what is needed to improve performance moving forward. Continuous improvement initiatives are beneficial, especially once all is up and running smoothly. With these initiatives in place, manufacturers can optimise processes and increase operational safety and efficiency. To keep an ongoing competitive advantage, manufacturers can consult with their trusted partner to help incorporate these initiatives as part of the overall long-term plan.

3. How is the relationship with the existing OEM? 4. Does the current equipment allow for modern security?

5. Does the existing system have the modern amenities for ease of maintenance and engineering?

6. Does the existing system have the modern amenities for ease of future innovations?

7. Does the existing system have the necessary memory and space for expandability?

8. Is the legacy system so old and antiquated that it would be easier to start over, especially if increased output could increase profitability?

9. Have the risks for both migration paths been identified, and a cost put to them?

10. From a conceptual perspective (FEL1 ±50%), which would cost more – piecemeal or rip and replace – and why?

Ten migration questions

A trusted, unbiased third-party partner can address all these factors, offer professional guidance and help brainstorm the pros and cons of an upgrade or migration. plus-circle

To help choose a DCS migration path, consider the following ten questions to help steer the overall project: 1. Is there evidence new technology could improve output?

2. Is there an existing installation base of new technology within the site that could be built or expanded upon?

Lynn Njaa is a senior consultant for Maverick Technologies’ DCSNext solution.

This article originally appeared on www.controleng.com

JEDER SPRICHT ÜBER DAS IIOT … wir setzen es einfach um.

Netzwerke und Computer für eine „smartere“ Industrie.

Leistungsstarke Computer für Ihre Bedürfnisse designt Sichere und verlässliche Netzwerke – immer und überall Vertikale Integration von SCADA bis zu Feldgeräten Moxa. Wo Innovation passiert. www.moxa.com

CYBERSECURITY

Myth busting Alvis Chen shares some best practice advice to help enhance industrial network security.

C

ybersecurity is a major concern for those deploying IIoT or Industry 4.0 systems. Often, when operational technology (OT) systems are connected to the Internet or connected to other IT systems, they become a point of weakness for malicious attacks or accidental data loss. But why is cybersecurity so often overlooked by OT engineers?

The answer can be traced to four common myths – misconceptions which are no longer true in today’s highly interconnected world Myth 1 – My industrial network is physically isolated and not connected to the Internet, so my network is secure: This may have been the case 10 years ago but today many IIoT devices are already directly connected to the Internet, bypassing traditional IT security layers. A question that is often asked is: Why do so many IIoT devices need to connect to the Internet? The main reason is because IIoT systems need to collect large amounts of data for further analysis. Since the data sources may not be in the same locations, it is necessary to send the data to a remote server by connecting your systems to the Internet.

Even if your industrial control systems (ICS) or industrial networks are not connected to the Internet, they may still be vulnerable to unauthorised connections. For example, a third-party vendor or an automation engineer may update systems by connecting unauthorised laptops or USB drives to conduct regular maintenance or troubleshooting, which opens the ICS up to insecure access and ultimately makes ICS devices more vulnerable.

Myth 2 – Hackers do not understand ICS, PLCs, and SCADA systems, so my network is secure: Since 2010, there have been several sophisticated

12

April 2020

cyberattacks that targeted ICS networks.

There has also been malware designed to target industrial control devices.

This trend indicates that hackers are changing their focus to target industrial sectors, such as oil and gas, energy, and manufacturing, which suggests that attacks on industrial sectors are likely to increase in the future.

Myth 3 – My network is too small to be targeted, so my network is secure: Internal breaches often come from trusted users, employees, and external contractors that have authorised access on a network. Often times, the unintentional breach is due to human error or a device that malfunctions, which is not relevant to the size of the company. Although these attacks are unintentional, they can still result in substantial damage and financial losses to business.

Myth 4 – I already have a firewall to protect my industrial network, so my network is secure: Firewalls may provide the first level of protection but they are not 100% effective. Moreover, most firewalls are not designed for industrial protocols (for example, Modbus TCP, EtherNet/IP, and PROFINET), so without proper configuration, the firewall may block necessary industrial protocols and shut down industrial control systems. Simply put, implementing firewalls cannot guarantee complete protection for ICS networks. Instead, industrial firewalls should be utilised with layered defences (the defensein-depth approach) to protect critical control devices, production lines, and the entire factory. In addition, industrial devices should be frequently updated with security patches to protect against cyberattacks.

Network security best practice Despite the differences in priorities and techniques used to protect industrial

www.controlengeurope.com

control systems versus enterprise IT systems, several industrial associations have developed standards and security guidelines for connecting or converging ICS with IT systems. In particular, the Industrial Internet Consortium (IIC), National Institute of Standards and Technology (NIST), and International Electrotechnical Commission (IEC) focus on three major areas for improving ICS cybersecurity.

The three pillars for securing industrial networks include: • Deploy defense-in-depth protection for industrial networks.

• Enable security settings on industrial networks.

• Manage security through education, policies, and monitoring.

Based on these three pillars the following best practices are recommended as the first step to shoring up an ICS cybersecurity.

Pillar I: Secure network infrastructure.

Secure networks are made by design.

Unfortunately, most automation networks have been deployed, added to, and modified slowly over years or even decades. Many PLC networks and devices were not designed to be connected to a plant network or the Internet and often lack strong security features. Since the priority was to keep the plant operating, networks were designed more with simplicity in mind than security.

In order to deploy a secure industrial network, the first thing that must be considered is a ‘defensein-depth’ network design. This will start with segmenting the network into logical zones, each of which is isolated and protected by industrial firewalls. Between each zone, set up conduits, which are firewall rules that filter or manage data communication across the zones in the network. In short, a defense-in-depth design seeks to protect the network from the inside out.

Consider the example of a smart Control Engineering Europe

CYBERSECURITY factory. Although it is important to deploy a firewall between the IT network and the OT network, this is not enough. Within the OT network, additional firewalls for critical assets, such as a controller for a distributed control system (DCS), should also be installed. Making it harder for unauthorised personnel to access a critical system minimises the potential impact of a security breach by limiting access to a single zone rather than granting complete access to the entire network.

An intrusion prevention system (IPS) or intrusion detection system (IDS) is an advanced system for industrial networks. The IPS/IDS will monitor network data for malicious activity. It is commonly used in IT/office networks, but can also be used for ICS networks as there are many applications that run on Windows-based industrial computers.

Another important factor in secure network design is secure remote access.

Similar to using VPN software on a laptop to access a corporate network from home, it is also possible to deploy encrypted VPN connections for remote monitoring or remote maintenance.

that have a well-defined response plan for patching vulnerabilities is more important than ever.

Pillar II: Hardened device security.

Another best practice for shoring up industrial network security is device security – often referred to as device hardening. This refers to securing the network switches, routers, and other devices connected to the industrial control system. Some of the methods include user authentication, maintaining the integrity and confidentiality of data, and using authentication to control network access.

While these concepts will be familiar, it is quite common to see industrial devices in critical systems deployed with little to no configuration for security.

Besides the previously mentioned security settings, also consider vulnerability management. Because vulnerabilities can affect components from virtually every software and device manufacturer, working with vendors

Pillar III: Security management and education.

The third best practice is the concept of security management or monitoring network security, which includes educating/training engineers that use the ICS to comply with new security policies. Education to ensure cybersecurity policies and practices are followed through could be the most important best practice of all, as well as the most difficult to implement successfully. To facilitate compliance, also consider investing in specialised software tools to manage ICS security policies more efficiently.

In particular, industrial network management software can help scan network devices, give an inventory list to allow for easy identification if something that should not be there is located, and remove it. Some tools can even help consistently configure

Control Engineering Europe

Firewalls for IC network should utilise layered defences to protect critical control devices.

www.controlengeurope.com

new devices to comply with the selected security settings, visually validate that the devices have been properly configured, and even back up configuration files to aid in network recovery if an incident occurs.

Another important feature is real-time event notification and logging. Logging can help pinpoint vulnerabilities and fix them before damage is done. Security information and event management (SIEM) systems are very important components in IT network management. Consequently, some industrial network management systems also offer APIs (for example, RESTful APIs) or support for common network protocols (for example, SNMP) for ICS integration with existing SIEM systems. plus-circle A Moxa Whitepaper entitled ‘Industrial Network Cybersecurity: Debunking the myths and adopting best practice’ can be downloaded from: bit.ly/2IqEVM0 Alvis Chen is a project manager at Moxa.

April 2020

13

CYBERSECURITY

HOPE FOR THE BEST BUT ALWAYS PREPARE FOR THE WORST Prof. Tobias Heer and Lukas Wüsteney highlight the relationship between preventive measures and possible responses that can be applied in industrial networks so that enterprises are better equipped to handle cyber attacks in the future.

W

ith security incidents at major companies are on the rise, trust in IT security has been fundamentally shaken. Taking a closer look at the current uncertainty, however, a new area of responsibility emerges: If we assume that we will be hacked at some point the question arises as to which measures and mechanisms have been put in place to adequately handle an attack.

To be prepared for an attack and to mitigate the damage or extent of that attack, it is important to start with a tried and tested concept – segmenting the network infrastructure and enhancing its security. In many documented cases, automated malware and attackers were able to move through networks systematically, as they were not separated well enough, or they contained too few firewalls to restrict an attacker. Many systems were therefore open to further attacks or the sabotage and disruption of production lines. Every industry network should separate the IT network from the OT network, as well as any functionally independent components of the latter.

Network segmentation follows the principle of zones and conduits.

The network defines functionally independent zones which, for the most part, operate autonomously. Packet filters (firewalls or gateways) are installed between the zones (see Fig 1), which limit and monitor network traffic that still has to flow over the borders of the zone. An industry network with strong segmentation not only contains a firewall between the OT and IT networks, but also a

14

April 2020

plethora of firewalls between individual system components and machines. Soft targets (old network devices that can no longer receive patch updates or do not have adequate security features) are separated from the network using packet filters, e.g. small firewalls that conceal the full scale of potentially vulnerable interfaces from attackers.

Zoning makes it more difficult for an attacker who has already infiltrated the network to move around freely.

Zoning is also an impediment for human hackers, as only the assets within the attacker’s zone are potential targets. A cyber attacker will generally try to move through the network to gain access to other computers. In that respect they depend on the existence of assets with vulnerabilities or inadequate configurations. Additionally, many system components retain functionality after an attack, as they are independent of the systems under attack. That means that outdated and unreliable measures, such as paper batch cards and manual work, need not be undertaken extensively in the aftermath of an attack since the attack is contained.

Cyber hygiene Generally, the first step for an attacker is to gain access to a system within the network. That can occur through infiltrating one of its own systems via an open, non-secure Ethernet port or a compromised Wi-Fi network. The attacker can also gain access through a system that has been compromised with malware and infected by a download or email.

In the event of the former, it is easy to recognise that an unknown asset

www.controlengeurope.com

(that of the attacker) is present on the network. An alarm can be triggered that allows for a rapid response to this deviation from the norm.

However, it is much better if the asset does not gain access to the network in the first place. Protocols such as IEEE 802.1X for Ethernet, and WPA2Enterprise with IEEE 802.1X for Wi-Fi, allow unique access credentials to be specified for each asset on the network.

Each asset wishing to access the network must provide authentication before it can communicate with it. This makes it more difficult for an attacker who has gained access to the company or production system to do the same with the network.

In reality, cases where industry equipment does not support authentication via 802.1X or WPA2Enterprise continue to crop up. In such cases with wired connections, a method known as a MAC bypass must be used, whereby the asset is recognised solely by its MAC address. Wi-Fi without IEEE 802.1X uses a WPA2 with pre-shared key, well known from home networks, where all Wi-Fi devices connected in this way share one secret key, the ‘WiFi password’. In such cases, however, it is no longer possible to determine absolutely which device wants to connect to the network, which is why additional measures must be taken to maintain network hygiene. Such automatic processes are often available as a post-connect phase (see Fig 2) in modern Network Access Control (NAC) solutions. If an attacker has already compromised a system within the network it is considered a legitimate network asset, meaning measures such Control Engineering Europe

CYBERSECURITY

Fig 1: Network segmentation and control systems.

as IEEE 802.1X and WPA2-Enterprise will be of no use. Additional methods must be used to monitor the state and integrity of the asset. Unlike checking its identity prior to connecting an asset to the network, an automated system can interact with the asset after connection (post-connect checks). This is achieved by either installing an agent (a small piece of software) on the system, which can identify itself and monitor critical system properties, or by making use of an alternative solution that does not use an agent. In the latter case, a control system notifies the monitored asset and subsequently uses commandline instructions to check the state of the asset. The parameters include the state of the anti-virus system, the state of the assets’ firewall, programs started on the asset, available or non-available files and registered users. This allows the typical behaviours of an attacker to be recognised. Assets that are known to be insecure are subsequently removed from the network automatically.

This makes it more challenging for an attacker to further infiltrate the network after it has taken control of an asset and remain hidden. It also allows for initially successful attacks Control Engineering Europe

to be detected and mitigated sooner.

Actively monitoring the state of the assets connected to the network and having effective network access control are therefore a crucial element of cyber hygiene in industrial networks.

A network access control management solution can also support asset management. The knowledge of all assets and their functions is an important basis for risk analysis prior to an attack and for deciding which response measures should or should not be taken (switch off, isolation, keep it running) in the event of an attack.

Preparing for emergencies It is pointless having a defence plan, if it does not include measures for detecting attacks. A technically superior attacker will otherwise be able to circumvent the defence mechanisms and move around the entire network, carrying out as many malicious activities as possible.

The concept of technical superiority sounds much more sophisticated than it is in reality. Consider that in 2019 one-quarter of all cyber attacks could still be attributed to the ‘Wannacry’ ransomware that has been causing malicious activity since 2017 and for

www.controlengeurope.com

which patches have also existed since 2017.

Security Information and Event Management Systems (SIEM) are used to recognise and classify cyber attacks in log files. Today, they are standard in many office IT environments, but are still rare in many industrial networks.

Despite this, the necessity of SIEM systems in industrial networks is just as important as in IT networks because the technology used, and the tools and methods used by attackers, are often similar or even identical.

However, by compromising devices, attackers inevitably leave behind traces.

For example, telltale registry entries, attack programs and files remain, or suspicious processes have been started on the systems or system configurations such as anti-virus and firewall settings have been changed. These can all be retrieved from the event logs of the compromised devices. If you view the logs from all the assets, you will get a general picture of all the activities going on in an industrial network.

The event logs and traffic pattern analyses contain valuable security information but this is often not exploited as they can be difficult to

> p16

April 2020

15

CYBERSECURITY handle. For one, it is not affordable to centrally collate information from hundreds of assets without a suitable management system (log management system). The other side of the coin is that it consists of a plethora of very small pieces of information from which it is not easy to identify an attack.

In addition to offering the potential to analyse huge quantities of log data, SIEM systems also allow for the automated recognition of attackers.

These systems are known as intrusion detection systems (IDS). Not only do they analyse log data, but also network traffic between the assets in an industrial network.

Once an attack has been recognised, countermeasures are automated to mitigate it. These intrusion prevention systems (IPS) can interrupt partially successful attacks and prevent further malicious activity. They have a bad reputation in industrial systems because if they are used without consideration they can lead to breakdowns, downtime and functionality errors which are tricky to fix. This bad reputation means that IPS systems have, in the past, responded with overly aggressive blockades in the event of an attack. For security purposes, whole network segments or assets were separated from the network in the event of a wrongfully recognised attack. This can result in extensive functional downtime.

Today, it is possible to use IDS systems more systematically, so that even in the event of an attack, a control asset can continue its tasks, while the configuration interface of interest to the attacker can be separated from the network by the IPS system. Through the isolation of such components, essential assets can continue to operate even in the event of an attack while the spread of the attack can be curtailed.

After an attack Finally, rapid and effective responses are required in the event of a known attack. To be ready to respond to an attack, full and detailed data for

16

April 2020

Fig 2: The attacker is limited to a single zone.

analysis must be stored in the log management system and trained employees must be available. The log management and SIEM systems mentioned previously are important here as they make information available in a format that is easy and efficient to search. This is key as a central logging system will often contain gigabytes of log data, meaning that recognising an attack is not straightforward. On account of the enormous volume of data, the entire process is challenging from a technological perspective too.

If you establish an effective system to handle emergencies on time, you can respond rapidly in the event of an attack. Without such precautions, in the worst-case scenario there is a risk that many of the affected systems could be separated from the network for a long period of time and that log data might have to be manually copied onto USB sticks for analysis.

In addition to technical systems, the capabilities of workers also play a decisive factor in the ability to actively fend off attacks or to restore systems afterwards. New employees should have qualifications in suitable technical fields and incident response plans must be continuously updated and properly communicated. Larger enterprises bundle their security competences into security operation centers (SOC). Additionally, advance contact should be made with

www.controlengeurope.com

specialised service providers to ensure additional support with incident response expertise in the event of an attack.

Emergency drills are also essential.

Statutory provisions regarding response times argue in favour of a team that is skilled in handling cyber incidents. For incidents where customer or employee data is affected, a notification period of 72 hours applies in Europe. For companies classed as critical infrastructure, the Federal Information Security contact persons must be notified immediately.

Considering these tight timelines and the working hours of employees, businesses often have no choice but to have IT security experts on site to handle emergencies.

Finally, it is essential to foster awareness and provide full training to the entire workforce in order to prevent attacks that use social engineering techniques such as phishing, deception, and baiting and it is important to provide training for less IT-oriented workers in administration and production if they have access to computer systems. plus-circle Prof. Tobias Heer is senior architect for Network Security at Hirschmann Automation and Control/Belden. Lukas Wüsteney is an architect for industrial networking at Hirschmann Automation and Control/Belden.

Control Engineering Europe

NEWS

REINVIGORATE THE WORKFORCE TO IMPROVE PRODUCTIVITY Manufacturers must focus on workforce wellbeing to help overcome the productivity crisis according to James Herbert.

P

roductivity in the UK is in decline.

A recent report from the Office for National Statistics shows that between June and April last year productivity dropped at its fastest annual pace in five years. And the worrying part is, it’s been flatlining since the economic downturn of 2008.

What can be done to turn this situation around? By exploring some of the common barriers to productivity, businesses across the UK can identify and implement solutions that will help to reinvigorate the workforce and improve performance.

Reducing absenteeism is a key objective for HR managers. However, too much time at work and working long hours can also impact productivity in a negative way. This is known as presenteeism.

Presenteeism should be tackled with the same vigour that businesses address absenteeism. Encouraging or allowing workers to continually work longer hours, or come to work when they’re suffering from poor physical or mental health, will inevitably lead to poor productivity.

Employers need to ensure their workers are striking the right balance. But both absenteeism and presenteeism can be exacerbated by the next barrier in this article…

Financial stress Financial wellbeing is something business leaders probably don’t think about when they think about the barriers to productivity. However, research shows financial stress is a widespread issue in UK workforces – and it directly impacts productivity.

Hastee’s Workplace Wellbeing Study 2018 found that one-fifth of workers admit to wasting working hours dealing with repayments. In 2019’s instalment of the study, respondents confirmed that financial stress impacts their work, as well as their health, sleep, social lives and relationships – all factors that can undoubtedly impact their performance at work.

The 2019 study found that only 21% of workers say they are able to budget and live within their means.

It’s clear that borrowing to get by is rife within UK businesses. Not only is this distracting people at work – in some cases it has prevented them from getting to work with 39% admitting they have been unable to make it to work due to financial difficulties.

Financial stress doesn’t just affect those directly coping with it. It’s impact on behaviour can create a demotivating environment for the

people around them. This brings us back full circle to absenteeism and presenteeism. A poor working environment can lead to people taking more sick days. On the other hand, if they’re dealing with a work environment that makes it difficult to concentrate, they might slow down or become increasingly distracted from the task at hand.

An ageing workforce It’s no secret that the manufacturing sector relies on a workforce that is predominantly older. Over-40s typically make up the majority of the labour pool and this is one of the highest concerns for 75% of manufacturers according to a recent YouGov survey.

The issue isn’t just that the workforce could be growing tired, with many approaching retirement age. Manufacturers are also struggling to attract new blood to the sector. The British Chambers of Commerce (BCC) reports that more than four-fifths of manufacturers struggled to hire the right staff in the final months of 2018.

Recruiting younger workers to a sector, that is often wrongly-viewed as ‘unexciting’, requires more exposure to the opportunities the sector provides.

Manufacturers also need to align themselves with destination employers that attract younger workers through the benefits and culture they offer.

By implementing benefits that help workers tackle issues such as financial wellbeing, employers can attract younger workers and ensure they are living relatively stress free lives which will support their workplace producitivity. plus-circle James Herbert is CEO at Hastee.

Control Engineering UK

www.controlengeurope.com

April 2020

UK1

WIRELESS SENSORS

ENERGY HARVESTING FOR SENSORS Gathering data from wireless sensors is critical in the Industrial Internet of Things (IIoT) era. Various energy harvesting methods can provide power. Chris Vavra reports.

A

well-planned Industrial Internet of Things (IIoT) infrastructure can allow a small team of workers to do the work of dozens more by leveraging data from thousands of devices on the plant floor.

This might seem daunting, but Will Zell, CEO and co-founder of Nikola Labs, sees this as the convergence of data and technology advances. Now, more than ever, workers have the ability to get a greater picture thanks to sensors attached to the devices on the plant floor.

“These sensors can provide data each day and see a picture that is being painted of what is and isn’t working in the plant,” Zell said in his presentation.

‘Wired versus Wireless for IIoT Solutions: The Pros, Cons, and Key Considerations,’ at Fabtech 2019 in McCormick Place in Chicago. “By leveraging sensor-based technology, you create a multitude of value.”

This is critical because unplanned downtime is costing manufacturers hundreds of billions of dollars each year. While companies are moving from reactive maintenance to preventive maintenance, Zell said the cultural shift can really take hold with the IIoT and improve automation and plant safety.

Some fear automation might be replacing workers on the manufacturing floor. Zell sees it differently. “The real challenge is how you change a maintenance team by installing sensors that can help the team without replacing them. Make them feel proactive. If this is executed well, a small team can leverage thousands of machines to do the work of dozens of humans,” he said.

Wired versus wireless Sensors, regardless of type, are designed to provide information, including how

UK2

April 2020

motors and drives operate. The most common challenge is the sensors have to be retrofitted onto the device because very few are installed with built-in technology.

The equipment involved may be 20- to 30-years old.

Wired industrial Through energy harvesting IIoT can deliver maintenance-free wireless sensors have sensors through sources such as vibration, RF and light energy. continuous Image courtesy of Chris Vavra. power and data reliability. Users don’t have to worry energy requirement than a pressure or about transmitted information getting temperature sensor. Zell believes that lost. The upfront costs for a wired sensor properly managing harvested energy, are high, and the sensors may be located sensing and data communication is key in areas difficult to access. to achieving a maintenance-free wireless Wireless sensors have emerged through sensor. He said: “Sensor platforms that the IIoT and are easier and cheaper have energy harvesting that can deliver to install compared to wired sensors. data to IIoT systems.”

Wireless sensors have more limited data Zell provided an example of RF capture and require batteries. Battery life wireless powering from a sensor over has increased, but batteries do need to distance via RF signals transmitted from a be replaced. For a large facility, that could transponder. From there, an RF harvesting cause all kinds of problems for the plant chip receives the signal to regulated dc team because thousands of sensors may power. need to be brought back up to speed.

According to Zell, delivering IIoT at scale can be achieved with wireless Wireless powering energy harvesting. While there is a The problem can be alleviated through place for wired sensors, particularly ambient energy harvesting and radio for applications that cannot afford frequency (RF) wireless power delivery. downtime, the overall benefits of energy With energy harvesting, Zell said, the IIoT harvesting will provide more reliable can deliver maintenance-free wireless data at faster speeds at lower costs for sensors using vibration, RF and light manufacturers. plus-circle energy. How much energy depends on the data payload and sensor type. An Chris Vavra is associate editor, Control ultrasound or video has a much different Engineering, CFE Media and Technology.

www.controlengeurope.com

Control Engineering UK

Game Changer Two Steps Ahead

ctrlX Automation from Rexroth: The most open automation system on the market Enter the future today with ctrlX Automation, a system with endless possibilities. This 360 solution is uncompromisingly open, with no proprietary systems or interfaces, free choice of programming language and ready for future standards. Flexible, with straight forward start up and nonstop performance, ctrlX keeps you entirely connected to almost everything. ctrlX, the revolution in engineering. Two Steps Ahead.

NEW PRODUCTS

Better data collection from hazardous areas SKF has gained hazardous area approval for its QuickCollect sensor, which collects vibration and temperature data. This new approval allows the sensor to be used in places that would previously have required a ‘hot work’ permit.

The certification opens up condition monitoring applications in a range of sectors, including petrochemical, mining and marine industries. It also brings digitalisation of data into hazardous areas, which up to now is a relative rarity.

The sensor is certified to both the

international IECEx and European ATEX standards, for use in Zone 1 hazardous areas.

The QuickCollect sensor can be used in conjunction with SKF’s ProCollect mobile app which connects the sensor to SKF’s web-based software platform, Enlight Centre. Together, this creates a portable condition monitoring system, called SKF Enlight ProCollect, that offers direct access to SKF’s experts in remote diagnostic services.

The package is available as a subscription service, which is charged

monthly or annually. This model allows SKF to tailor the package to the customer’s exact needs.

Non-invasive inspection of oil and gas separators An optical gas imager from FLIR Systems is said to be safely and efficiently enabling oil and gas producers to undertake routine, non-invasive inspections in their oil and gas separators.

Separators are pressure vessels designed to separate a well stream into gaseous and liquid components. During this process, sand can build up in the heating element, causing damage to the separator. If this build-up is not noticed and promptly removed, it can result in the emission of flammable gases, costly repairs or catastrophic failure.

Traditionally, oil and gas separators have been inspected by hand or cleaned out on a routine basis without evaluation. While this process provides basic maintenance, it does not provide the early detection required to identify mechanical or safety issues.

A variety of devices can be used to inspect oil and gas separators to verify tank pressures and liquid levels however, use of thermal imagers are one of the safest and most efficient solutions. The FLIR GFx320 optical gas imager can be used to visualise natural gas leaks and sand levels as the camera shows changes

in temperature from sand versus oil, gas, and water.

The GFx320 is certified for use in Class 1: Division 2 or Zone 2 locations, allowing the user to get close to the unit for a thorough, but non-invasive inspection.

The use of the GFx320 for routine inspection of oil and gas separators is allowing tank pressures and liquid levels to be verified. It can also detect leaks from the separator, resolve anomalies to restore compliance, avoid adverse environmental impact, and prevent safety risks to equipment or personnel.

ATEX approval for ultrasonic clamp-on flow meter Badger has announced new hazardous area certification to ATEX/IECEx Zone 2 for the Dynasonics TFX-5000 transit time ultrasonic flow meter.

The certification covers models measuring DN65 and larger, making the solution suitable for flow measurement in automotive, waste-water, oil and gas and other general industrial applications requiring hazardous zone approvals.

The ultrasonic clamp-on flow and energy meters are said to offer a versatile solution for measuring volumetric flow and heating/cooling rates in clean liquids as well as those with small amounts of suspended solids or aeration, such as surface water

UK4

April 2020

or raw sewage. Typical applications include water mains, reclaimed water, lift stations and booster pump stations in water and wastewater or the energy

www.controlengeurope.com

transfer of chilled water (cooling) and glycol/hot water (heating) in HVAC systems as well as produced water in oil and gas applications.

The TFX-5000 meter is intended for users seeking an accurate, reliable and affordable transit time metering device. It is available in a variety of configurations and can be selected with features suitable to meet particular application requirements.

By clamping onto the outside of pipes, ultrasonic devices do not contact the internal liquid which offers inherent advantages such as reduced installation time and cost, no limits on pressure or fluid compatibility.

Control Engineering UK

CYBERSECURITY

Stay alert to your vulnerabilities David Emm, principal security researcher at Kaspersky, answers questions about industrial cyber threats and vulnerabilities.

Q: What is the current threat landscape for manufacturing companies?

Industrial and manufacturing cyberincidents have the capability of shutting down entire countries or creating tangible financial loss, as we saw following an attack on the Ukraine power grid in 2015. It has been estimated that during the first half of 2019, more than one-third (41%) of industrial control systems (ICS) were victim to cyberattacks, triggering security software to step in and remediate. Infection with such malware can negatively affect the availability and integrity of ICS and other systems that are part of an industrial network, making it difficult to resolve.

The manufacturing industry is already a known target for attackers, and therefore the most common threats of spyware, cryptocurrency miners and worms still occur.

Q: Where are the primary threats for manufacturers – internal or external?

Research has shown that the greatest danger to manufacturers is in the form of cryptocurrency miners (3%), worms (7%) and versatile spyware (4%). These types of attacks can come from internal or external sources, and more often than not, attackers will target employees using the system with phishing websites in order to get access to company data.

Security experts and IT departments should be particularly cautious about malicious software that aims to steal data, spy on critically important objects, penetrate the perimeter and destroy data.

Q: What are the main areas of vulnerability for manufacturers?

A cybersecurity incident that occurs because of a targeted attack or infection of conventional malware can lead to damaging consequences and a disruption in manufacturing processes.

Control Engineering Europe

For many industrial or manufacturing organisations, while they actively invest in the cybersecurity of corporate networks, cybersecurity in OT/ICS networks and securing the OT environment can be an afterthought. As with any organisation, if you don’t take steps to protect yourself, your employees and your business as a whole, then you become an easy target for cybercriminals. You are essentially wide open to attack; so, failing to recognise the threat cyberattacks can impose on a business can have an extremely damaging outcome. Criminals actively look to sabotage computerised systems to impact the delivery of services the company is supposed to provide. This can result not only in lost data and a damaged reputation, but can also have a huge financial impact.

Q: Is there a one-size fits all cybersecurity architecture?

When it comes to managing cyberthreats, businesses need to be prepared. It is important to regularly update operating systems, application software and security systems that are part of the enterprise’s industrial network. Restricting network traffic on ports and protocols used on edge routers and inside the organisation’s OT networks is also vital, to stop the wrong people accessing an organisation’s data. But it is not only the infrastructure that should be prepared for attacks; employers should provide regular training for staff, to allow them to spot the difference between a genuine email or website and a phishing attack.

www.controlengeurope.com

Q: What new technology is available for manufacturers to protect against attacks?

Technological development has made a huge difference to the way cyberattacks are identified and the speed at which this can happen. Artificial Intelligence (AI) and machine learning have meant that we are able to keep up with attacks and stay ahead of them. The systems allow us to process one million potential attacks a day – something that could never happen if we didn’t automate our processes. This is the expertise that feeds into different levels of protection, from endpoint to anti-targeted attacks and threat intelligence, to ensure businesses are secure at all levels and attacks can be prevented.

Q: If you were to give one piece of advice to manufacturers on cybersecurity what would it be?

It is difficult to call out one single thing, but I would advise any organisation – be it a manufacturer, business owner or an individual – is to ensure your operating systems are up to date. Cybercriminals make use of the vulnerabilities that occur when systems are behind on updates, and this can leave organisations wide open to attack. plus-circle April 2020

17

MACHINE VISION

HAVING A VISION FOR AI AND DEEP LEARNING Advances in deep learning/AI is resulting in these technologies being increasingly utilised within machine vision solutions. Control Engineering Europe sought advice about how end users can ensure that they are able to implement successful AI-based machine vision applications

N

eil Sandhu, UK product manager for Imaging, Measurement & Ranging at SICK, believes that AI/ deep learning machine vision will result in greater production flexibility because it has the potential to retrain machines, adapt to changes in processes and respond to a high variety of products – all of which are, of course key elements of Industry 4.0.

“Deep Learning technologies should be especially attractive to end users because they can cut out tedious and lengthy programming time and costs especially for more complex tasks,” he said. “This offers the potential to automate machine vision tasks that have previously been too difficult, costly or time-consuming.”

However, Sandhu goes on to warn that deep learning should not be considered as a silver bullet for every application. He believes that it is suited to harder-to-solve inspections where there are a greater number of natural variations from a standard, which would be laborious or even impossible to solve one at a time.

Ruben Ferraz, field product marketing manager Deep Learning at Cognex, pointed out that, as with any new technology there are considerations and trade-offs so the advice is to set proper expectations for what deep learning can bring to any project. “It is important to understand these trade-offs at the outset,” he said.

With any deep learning project, there are four core job roles that are needed for resource planning. These include:

18

April 2020

1. A vision developer who implements the solution, as well as optimises lighting and image formation; 2. A quality expert who analyses and grades images; 3. An image labeler; and 4. A data collector who records and organises all information including images, grades, labels, and meta data.

While it is possible for one employee to cover more than one of these roles, being aware of the types of expertise needed is helpful to have upfront.

It is also worth nothing that any deep learning initiative will require a powerful Windows-based PC with a graphical processing unit (GPU) installed.

Ferraz advises that the best route forward is to pilot small manageable projects in a sensible phased approach to allow automation teams to set themselves up for long-term success with deep learning image analysis.

“Pick a project with a clear payback that cannot easily be solved with traditional rule-based vision, but which is not so difficult that it never makes it into production. Focus on a core need and develop both a core competency and understanding of what deep learning can and cannot do in a factory automation setting. Deep learning pilot projects should have two primary goals: evaluating its broader utility for a more holistic automation strategy

www.controlengeurope.com

and automating an inspection or

verification process that is either not done at all or done manually,” he said.

Change is coming Due to complexity, variability, and the necessity to distinguish between very small differences some inspection applications have been impossible to achieve with traditional machine vision systems. However, things are now changing, according to Damir Dolar, director of embedded engineering at FRAMOS. He said: “Deep learning can Control Engineering Europe

MACHINE VISION cope with complexity and variability.

It combines the flexibility of human inspections with the consistency and speed of a computer. Experience accumulated in the past can now be applied and used to train deep learning systems.

“The ability of AI to process images is well-documented and most manufacturers will already have large databases of past material that could easily be used by deep learning algorithms for initial training. Once the algorithm has been trained, it can be deployed, maintained, and improved over time just like

“In very simple terms, a deep learning algorithm could easily automate quality control by learning what a ‘good’ part looks like and rejecting the rest,” he continued. “More broadly, and if it is not limited to the specific visual inspection system, deep learning algorithms could be used to learn what part of the manufacturing process causes a set of defects. The algorithm can adjust the current process and remove all defects from subsequent issues, they are able to locate parts despite a change in appearance, distinguishing between functional, and cosmetic defects. It also detects classifications while tolerating differences in size, appearances, orientation, and naturally occurring variations.

“The main advantage of such an application is in consistency of product quality, because the algorithm operates 24/7 and maintains the same level of quality all times. The system identifies every defect out of tolerance, the operation is faster, defects are identified in seconds, and it supports high-speed applications.”

The best approach

any other piece of industrial equipment.”

According to Dolar AI and deep learning is a useful tool in every machine vision application which currently has limitations of performance.

With deep learning it is possible to easily develop high performance solutions for difficult vision problems. Including, for example, automated quality control for goods and products that could not be monitored with current technologies.

Control Engineering Europe

Dr Robert-Alexander Windberger, an AI specialist at IDS, believes that deep learning/AI machine vision solutions are best approached iteratively and empirically. “It can be hard, or even impossible, to predict the accuracy of a neural network analytically, or the amount and quality of data required,” he said. “However, one of the big advantages of deep learning is that there is no need to cover the full complexity of a scenario by sets of rules but to improve the model by example online. While the evaluation of deep learning models on data sets helps to ensure a model's convergence and to benchmark its accuracy, many issues only arise in a physical test setup. So, the goal is to get a first model working in an application-related setup as early as possible.”

The exact starting point and the iteration step size will depend on the user's experience. Windberger advises

www.controlengeurope.com

that when in doubt it is best to reduce the problem to an absolute minimum: In a quality assurance scenario this could be a binary statement, ‘IO’ and ‘NIO’, instead of covering all occurring errors and locating them at once. “This may yield a fair gain if a major workload can be lifted from a manual quality control.

Further classes can be subsequently added in a controlled fashion – one at a time – and immediately tested.

“The reduced complexity requires only a reduced data set size accordingly, which can again be extended step-bystep. It is possible that data biases will become evident during the evaluation and it would be unfortunate if this happened at the end of an expensive data acquisition campaign. By having an application-related setup in place early in the process, data acquisition will be simplified. Through frequent evaluation, data can be obtained more purposefully: If certain image contents are found to be misclassified or overlooked, it might help to add these contents to the data set for retraining,” continued Windberger.

He believes that deep learning can deliver a specified result for certification but says it can also be implemented as a continuous process. “The solution can be steadily improved and adapted by topic experts, who use their knowledge to feed the algorithm handpicked data.”

Yonatan Hyatt, CTO and co-founder of Inspekto, says that AI and deep learning can drive and hyper optimise hardware to the specific need at hand.

He said: “Instead of using AI as a software stack in a larger scope project with many hardware layers, I would urge end-users to demand products that already include the required hardware and AI-based smart software. This will save time – both in the project planning phase and later in the operative stage – allowing for self-adjustment instead of repetitive iteration stages with the expert integrator.”

New application areas Control Engineering Europe then went on to ask what type of new machine vision applications we can expect to see April 2020

19

MACHINE VISION when AI and deep learning are properly utilised. Sandhu said: “Machine vision has always operated by matching examples, but deep learning takes this to the next level by using the knowledge of hundreds of experiences to deliver a judgement that classifies a product, for example as belonging to a ‘good or bad’ category.

“Deep learning software is developed using artificial neural networks that learn by example, in the same way that humans do, to model what a good part should look like and what variations can be tolerated. There is no need to select from the conventional toolbox of algorithms used to identify defects, such as pattern finding or edge detection. Deep learning cameras can automatically detect, verify, classify and locate ‘trained’ objects or features by analysing the complete ‘taught-in’ image library. As a result, the solutions most likely to emerge with deep learning vision will be specific and customised on a case-by-case basis.

A SICK pilot project has been operating in the timber industry to optimise the cutting process in a sawmill based on deep learning recognition of the annual age rings and other features in the lumber, for example knots in the wood. Using the technology, the sawmill has been able to make the best use of each log and avoid waste, while improving the overall product quality.

In the automotive industry, the complexity of all the potential creases or flaws in a leather car seat during quality checking following ironing, provides a good example of the degree of variation of surface features which can be solved by deep learning vision. Here, systems can be trained on hundreds of images to make a judgement on whether each subsequent example is a pass or fail.

Qualitative problems Windberger believes that qualitative problems are more suited to a deep learning approach than quantitative problems. Explaining further, he said: “If a trained employee makes

20

April 2020

a qualitative decision within a few seconds just by looking at an object, chances are that deep learning might be worth considering. If this employee had to use a caliper to come to that decision, then other methods might be more promising.”

Windberger argues that, while these qualitative decisions may seem to be the simpler ones, the fact that many companies still have these decisions embedded as manual processes suggests that this is not the case. “These processes often involve repetitive, monotonous tasks and frequently suffer from varying quality standards over time or across company subsidiaries.” The reasons for these manual processes to persist can be, among others, a high object variety, as in medical or food industry, or undefinable backgrounds and environments, as in retail or logistics.

Moreover, decision criteria can be difficult to quantify and cover with a set of rules. For example, if the aesthetic appearance is relevant, such as in the wood industry. If deep

learning can be implemented in a continuous improvement cycle, it can be an adaptable tool which can also cope with time-varying objects, such as annual variations on fruits, vegetables, weeds, and pests in the agricultural industry.

“Where rule-based machine vision has not been attempted or has reached its limits, due to one of the above reasons, there is a high potential for deep learning algorithms to support employees and drive forward automation.” Windberger stresses that the different machine vision approaches should not be viewed as competition, but rather as complementry. Often a qualitative decision hampers the implementation of a quantitative image processing chain, such as picking the right measurement algorithm depending on the item under investigation or finding the right area of interest to apply the measurement to. In these cases, access to the full machine vision tool kit, including deep learning, allows for powerful combinations. plus-circle Control Engineering Europe

WIRELESS TECHNOLOGY

5G BENEFITS AND BARRIERS A whitepaper from Digital Catapult takes a look at the important role 5G is set to play in ensuring connectivity to enable digitalisation of enterprises.

It also looks at the current barriers to 5G deployment.

5

G’s promise of ultra-low latency, extremely fast data speeds and the ability to simultaneously connect a million devices per km² looks set to open up new opportunities to optimise manufacturing processes.

Early test cases have shown that 5G represents a step-change in connectivity for the manufacturing sector and its technologies go far beyond what mobile networks have previously offered. 5G offers a unique combination of features that can answer almost any set of connectivity requirements for specific or multiple industrial digital technology (IDT) use cases.

With latency of less than a millisecond – five times lower than 4G – 5G technology holds the key to real-time processes. Further, its high reliability means it can be used for mission critical operations – manufacturers will be able to manage multiple connectivity technologies, including legacy networks, through a single 5G network. It will also give the ability to run a private 5G network, or have control of a dedicated ‘network slice’ from a mobile

22

April 2020

network provider. Both options put manufacturers in greater control of their own connectivity, security and quality of service.

Advanced manufacturing 5G use cases can improve efficiency and safety as well as reduce downtime. For example, 5G will be needed for large-scale predictive maintenance and time-critical hazard detection or scale deployments of collaborative robotics.

5G use cases can be categorised into three clusters: 1. On-site and in-factory production optimisation.

2. Monitoring and management of goods across the supply chain.

3. Connected goods: product life cycle management (including end of life).

To date, manufacturing companies have focused their industrial digital technology plans on production and in-factory processes, which are often viewed as most business critical.

Opportunities to manage incoming and outgoing goods could, however, be transformative, giving visibility of

www.controlengeurope.com

the entire end-to-end supply chain for the first time. Connected products, following entry into service, meanwhile provide an opportunity to build new business models for the sector.

While 4G fuelled mass adoption of mobile Internet and digitised our social lives, 5G looks set to do the same for industrial business use cases. It is the first cellular network technology designed to address machine-type communications and meet the requirements for multiple industrial digital technology use cases, from high density of sensors to autonomous vehicles.

The difference between 4G and 5G goes far beyond increased bandwidth.

For example, 4G, with a latency of 50 milliseconds (ms), is not fast enough to deal with tasks requiring real-time communication, while 5G at 1ms will do so perfectly.

Telecoms equipment vendor Ericsson worked with Fraunhofer Institute on a 5G test and case study focused on improving process control and speeding up detection of manufacturing failures for high cost metal blades used in turbines, including jet engines. They estimate that 5G capabilities, including ultra-low latency, can deliver a decline in rework rates from 25% to 15% – a machine cost reduction of €3,600 per blade. This would equal an annual saving of €27 million for just one factory. Its technical capabilities are grouped into three areas: Enhanced mobile broadband (eMBB): Likely to be the first deployments of 5G technology, to address the large growth in mobile devices and demand for data with 10+ GBps bandwidth. eMBB enables services such as streaming of ultra high definition (UHD) video, intelligent analytics of large volumes of data using artificial intelligence/machine Control Engineering Europe

WIRELESS TECHNOLOGY learning, and training and assisted operations using augmented and/or virtual reality.

Massive machine-type communications (mMTC): Utilising sub 1GHz spectrum to deliver large scale machine to machine (M2M) communication, mMTC enables large scale Internet of Things (IoT) deployments and roll-outs of sensors on-site, across large distributed sites and the supply chain, as well as connectivity between manufacturers and their endcustomers.

Ultra-reliable low latency communications (URLLC): Driven by new use cases such as remote maintenance and monitoring, collaborative robots (cobots) and connected autonomous vehicles, URLLC will deliver ultra-fast mission critical connectivity. This will enable highly accurate and reliable real-time data that can be processed, analysed, visualised and actioned at scale, both on-site and across the various parts of the supply chain. This feature is crucial in a manufacturing process with extremely high tolerance requirements.

Close to one millisecond latency and very high bandwidth make it possible to control manufacturing machines in real-time, reducing costs and improving quality.

It should be noted, however, that these three capabilities work on the basis of trade offs. For example, to achieve ultra low latency, there may be a need to reduce the device density or the data speed.

A crucial change delivered by 5G lies in how the network is managed. Currently, networks need to be managed separately. Using 5G technology, a company could simultaneously manage different: • Types of access networks: For example, wired, wireless, optical, copper.

• Technologies: For example fieldbus, Ethernet, wireless.

• Protocols: For example real-time, best effort.

• Equipment products from different equipment: Vendors may otherwise be incompatible.

Control Engineering Europe

Addressing the challenges Unfortunately, to counter the benefits, there are also many challenges to 5G adoption in the manufacturing sector.

These include: • A lack of demonstrable cost-efficiency and return on investment. This is further complicated by the fact that connectivity is typically not part of manufacturing companies’ R&D plans, despite awareness that current connectivity does not meet their future requirements.

• Concerns around compatibility and interoperability of mobile networks when it comes to integration into existing industrial systems.

• A need for security.

• Lack of understanding of how 5G differs from other connectivity solutions.

• Cultural barriers to working with companies in different sectors such as telecommunications, as well as startups.

Barriers to deployment A manufacturing survey, conducted in the UK by Digital Catapult, the UK’s innovation centre for advanced digital technology adoption, showed that 71% of respondents believe 5G will bring benefits to their organisation.

However, there are barriers to its successful deployment with each area of manufacturing – from aerospace and defence to fast-moving consumer goods (FMCG) – having its own sub-set of considerations.

Many emerging technologies are held back by a lack of certainty about the value they can bring. This often halts investment beyond a proof of concept or siloed use case. A clear ROI and business case is crucial to the introduction of 5G in manufacturing and it would appear that proven ROI in deployments are scarce.

The situation is further complicated by a lack of attention to connectivity at a strategic level within manufacturing companies. Most manufacturers see the value of automation and the introduction of sensors for efficiency

www.controlengeurope.com

and productivity. Several companies interviewed by Digital Catapult said their existing connectivity covers their current needs, but note that the wired connectivity typically used is inflexible and costly to expand to provide further capabilities. One manufacturer, for example, was looking to expand from 1,000 factory sensors to 100,000. This would be challenging, complex and possibly even physically impossible to do with wired connectivity, due to the cost of wiring, tray installations and disruption of production.

Manufacturers typically do not consider connectivity to be part of their strategy and R&D processes. It is often seen as a commodity and a cost centre.

Often, current connectivity solutions are assessed to evaluate performance as well as the cost of keeping, rather than replacing them, given efficiency, productivity and quality considerations.

The connectivity challenge is therefore often not part of considerations when undertaking proof of concepts using technologies such as IoT or machine learning. For a limited, siloed deployment, existing connectivity has been more than adequate.

There also appears to be concern about compatibility and interoperability with existing solutions. This is often driven by experience: for example, previous teething problems experienced when introducing new solutions at the heart of processes. If there are problems, consequences can be catastrophic. New connectivity solutions need to cater for existing standards in the areas of security and reliability. This includes specific regulatory requirements in many sectors. Concerns in this area mainly relate to personnel safety and end-to-end reliability of manufacturing processes.

Of course, security is paramount and 5G poses challenges here, due to its versatility. For example, enabling a large number of services and IoT devices increases the range of points potentially open to threat. An open, flexible, programmable network can also be more vulnerable.

> p24

April 2020

23

WIRELESS TECHNOLOGY Many manufacturers are also sceptical about having new outside parties – with limited knowledge of their operations – controlling their vital communications.

5G connectivity, based on network slicing provided by mobile network operators, could partly address the need for security, control and privacy. In this scenario, network operators or systems integrators could provide the advanced services required.

Research undertaken by Digital Catapult showed that manufacturers prefer a fully private network, enabling them to control the production line, giving them full flexibility over any modifications and ownership of risk and support.

The architecture of 5G, with virtual and disaggregated network functions and the introduction of edge computing, enables private networks.

Private network deployments and operations introduce new challenges for manufacturers, as these incur increased network infrastructure costs for both deployment and operations.

Manufacturers would also need teams skilled in cellular network deployment and management.

Completely private network solutions will also require access to radio spectrum. In the UK, for example, the regulator, Ofcom, has not set aside spectrum for use specifically in the industrial sector, but it is consulting on greater shared access to spectrum.

Private 5G systems also require new architectures. Several telecoms equipment vendors are working to create these. They aim to make networks as plug-and-play as possible, to remove the need for specialised telecoms engineers to manage them.

Lack of understanding Manufacturing companies recognise that their organisations are often not aware of what capabilities 5G can deliver. Only one-third of those surveyed

by Digital Catapult said they have a good knowledge of 5G, while 7% view themselves as experts.

Manufacturers often incorrectly

24

April 2020

believe that their existing connectivity capability already performs around 80% of what 5G is expected to deliver. This drives a reluctance to 5G for the UK investigate 5G, in manufacturing sector particular if there July 2019 is no confirmed business case.

Furthermore, connectivity solutions are often managed

by IT teams. The skill set in these roles does not extend to cellular connectivity in general, nor 5G in particular. This means in-house teams often do not have the knowledge to assess different options. and MHz. 5G involves the introduction So, for manufacturers to roll-out large of new systems which blur the line scale changes to processes based on between IT and connectivity. In addition, 5G, it will be imperative to re-train and new technology experimentation often upskill staff. involves collaboration with innovative For 5G networks to be widely startups who work, communicate and adopted, they will also need to be act in a very different way to large, significantly easier to build, configure, established manufacturing companies. commission and operate than they are today and this is an area where Conclusion companies adopting 5G and the For 5G to be deployed in a mobile industry need to collaborate manufacturing environment, experts and learn from each other. Until these in the various domains need to work capabilities are recognised as important together, using a common language. in the sector, there will be significant Third parties – with knowledge of all barriers to both digitalisation and sides of the 5G story – can help mediate the deployment of 5G. In a negative between the different businesses. circle, what often amounts to a lack of Systems integrators, who are already understanding of 5G also means it is trusted by the manufacturing industry, harder to determine its potential ROI. will also play a key role. Vitally, A language gap between the parties education and advisory activities will involved is also hindering strategic also be a key requirement. plus-circle discussions about the opportunity of 5G. Manufacturing engineers speak in terms The original Digital Catapult whitepaper of production, IT teams speak in terms document ‘Made in 5G’ of servers and cloud while telecoms can be downloaded from: providers speak about throughput https://bit.ly/2IyN94L

Made in 5G

www.controlengeurope.com

Control Engineering Europe

The future of Edge Computing is here

“

What you’re doing now are the early steps in a journey to get to where you’re ultimately going, which is edge computing.

- Jason Andersen, vice president, strategy and business line management Stratus Technologies

New research from Stratus Technologies and CFE Media shows that we are approaching the tipping point for edge adoption. The introduction of edge computing is the most significant plant-floor advancement since the introduction of SCADA 30 years ago.

Our new report can help you understand: • The barriers to implementing Edge Computing • The skills your teams need to optimise for Edge Computing • The top 5 Edge Computing applications

41%

see edge as a great leap forward

51%

are actively planning edge implementations

Download

YOUR Free Copy

TODAY

at

stratus.com/EUTrend

f

PREDICTIVE MAINTENANCE

TURNING THE DREAM INTO REALITY Philipp H. F. Wallner offers advice on how to realise the dream of implementing successful predictive maintenance applications.

P

redictive maintenance has been heralded as a solution to manufacturers’ and engineers’ woes thanks to its potential to give equipment users the ability to anticipate imminent malfunctions, proactively arrange repairs, reduce disruption to operations on the factory floor, and, most crucially, safeguard against the failure of equipment, which can have a disastrous impact on the whole business. Based on these use cases it is understandable why this technology is perceived to have huge value. According to Deloitte, predictive maintenance can reduce overall maintenance costs by between 5% and 10%.

But, the key word here is perceived value. While predictive maintenance has the capability to be truly transformative, when the time comes to implement the technology on existing equipment the process is not always straightforward and this fact is reflected in the actual number of businesses that have implemented predictive maintenance in operation, which is very few.

So why is the rate of adoption so slow? Industry commentators have highlighted four factors that seem to be the key stumbling blocks that need to be conquered by equipment builders and operators to effectively work alongside the data science community and realise the dream of successful predictive maintenance solutions. These factors include: 1. Encouraging teamwork and knowledge sharing to benefit from existing domain know-how during the algorithm design process: It can

26

April 2020

be difficult for businesses to cultivate a collaborative environment where powerful algorithms for predictive maintenance, based on statistics methods, are designed that integrate the domain knowledge and expertise of both data scientists and domain experts.

Furthermore, how can domain experts and data scientists work together to make sure that the key elements of each effective predictive maintenance application are fully leveraged? How can they be sure to include both data analytics methods and domain knowledge?

The best predictive maintenance applications will include both of these components: statistics-based data analytics methods, like machine learning, in addition to the domain expertise about equipment that the R&D engineers possess (very often already incorporated into existing simulation models). If predictive maintenance is approached with a singular data analytics mindset, users will not capture all of the useful information retained by the operations and engineering

teams that build the equipment and are responsible for their ongoing upkeep.

2. Determining how to train algorithms without access to sufficient failure data: Training an algorithm on data from the field is a fundamental part of machine learning. Those creating the algorithm must include ‘good’ data from everyday production on top of a variety of failure data taken from the numerous error scenarios that can happen while the equipment is operating. However, if the goal is to never allow the equipment to break in the first place, then where can the failure data be obtained?

This is turning out to be a progressively important problem to solve for businesses utilising predictive maintenance for their industrial systems. What’s more, it is irrespective of use cases and can range from air compressors to wind turbines. To overcome this issue, simulation models can be brought in to produce artificial failure data, so the algorithms have something to be trained on when there isn’t any, or not enough, measured failure data from the factory floor.

3. Taking the algorithms from the design stage to real-world operation: After the training and design of the predictive maintenance algorithm has been carried out on the desktop, the next step is deployment onto the equipment.

The difficulty level of this process directly correlates with the condition of the existing IT and OT infrastructure.

A simulation model of a triplex pump in Simulink and Simscape that is used for training a fault classification machine learning algorithm to overcome the absence of measured failure data. © 1984–2020 The MathWorks, Inc.

www.controlengeurope.com

Control Engineering Europe

PREDICTIVE MAINTENANCE

Whereas some algorithms are applied on a real-time hardware platform – for example, on an industrial PC, an embedded controller, or a PLC – there will be some that are in the cloud or will be merged with the current non-real-time infrastructure (for example, an edge device running on Linux or Windows).

At a growing rate, businesses are taking the option of implementing an efficient way of using toolchains that facilitate automatic generation of C, C++ or IEC 61131-3 code, .NET components, or standalone executables. For example, a manufacturer of packaging and paper products installed predictive maintenance software into its manufacturing line as a way to lower the amount of waste produced and reduce machine downtime in its plastics manufacturing facilities.

(See tint panel for details) 4. Proving the potential return on investment (ROI) of predictive maintenance solutions: When any organisation kicks off a predicative

maintenance project, the most important question it has to be able to answer at the outset is, how can I prove the ROI of this investment? Otherwise, it will struggle to get that initial budget going.

In the absence of an answer to this question, all the time and energy spent on developing a detailed predictive maintenance plan and solution will quickly run aground. Identifying a concrete business case and developing an approach for how to monetise predictive maintenance will prove vital when trying to persuade your corporation’s management team to rationalise the investment in executing a predictive maintenance project.

The most obvious quantifiable benefit for equipment operators will be the reduction in equipment failure during operation. While this often justifies the investment for operators, for equipment builders, building a case is more difficult.

However, there are a number of ideas that have been proposed by some of our clients, which have significantly

contributed to building a solid case for implementing predictive maintenance.

They can be summarised as: • Linking service fees to predictive maintenance of the equipment used by the operators – equipment builders’ customers.

• Taking advantage of IP protection to sell the deployed predictive maintenance algorithm itself.

• Moving to a new business model based on usage (for example, selling elevator usage hours rather than entire elevators, or cubic meters of compressed air rather than compressors).

It will be only a matter of time before the C-suite – armed with the information of these possibilities – jumps on board, realising predictive maintenance in all its glory. plus-circle Philipp H. F. Wallner is industry manager, Industrial Automation & Machinery at MathWorks.

A statistics-based health monitoring and predictive maintenance solution Mondi Gronau is a manufacturer of packaging and paper products – producing around 18 million tons of plastic and thin film products annually.

Machine failures that result in downtime and wasted raw materials cost the company millions of euros each month. To minimise these costs and maximise plant efficiency, it has developed a health monitoring and predictive maintenance application which uses advanced statistics and machine learning algorithms to identify potential issues with the machines, enabling workers to take corrective action and prevent serious problems.

“As a manufacturing company we don’t have data scientists with machine learning expertise, but MathWorks Control Engineering Europe

provided the tools and technical knowhow that enabled us to develop a production preventative maintenance system in a matter of months,” explained Dr. Michael Kohlert, head of information management and process automation at Mondi.

The machines used are large and complex. Each is controlled by up to five PLCs, which log temperature, pressure, velocity, and other performance parameters from the machine’s sensors. Each machine can record up to 400 parameter values every minute, generating seven gigabytes of data daily.

Mondi faced several challenges. The plant personnel had limited experience with statistical analysis and machine learning. The company needed to evaluate a variety of machine learning

www.controlengeurope.com

approaches to identify which produced the most accurate results for data. It also needed to develop an application that presented the results clearly and immediately to machine operators.

Lastly, it needed to package this application for continuous use in a production environment.

Mondi worked with MathWorks Consulting and Prof. Dr.-Ing. Andreas König to develop and deploy health monitoring and predictive maintenance software in MATLAB. A prototype was completed within six months and the result of the project has been a saving of over 50,000 euros every year. “This figure is based on just eight machines. We expect that to increase at least fourfold as we analyse the data from more of our machines,” concluded Dr Kohlert.

April 2020

27

INDUSTRIAL ETHERNET

BIG DATA ANALYTICS MOVES RIGHT TO THE EDGE Brendan O’Dowd looks at the implications of the introduction of the IEEE 802.3cg standard for industrial automation and what it might mean for 4-20mA or HART systems.

T

he ratification in November 2019 of the IEEE 802.3cg standard marked the introduction of a new and dramatically different way for factory operators to connect devices at the edge of the network, freeing them from the restrictions of infrastructure based on the legacy 4-20mA and HART communications interfaces.

The 802.3cg standard, also known as 10BASE-T1L, is a type of industrial Ethernet networking protocol. It provides a way to break down the barriers between the basic operational devices which perform frontline service in the factory or process plant – the sensors and valves, actuators and control switches – and enterprise data, where the intelligence of the new ‘smart factory’ comes to life.

10BASE-T1L networking looks set to become an important enabler of the general transformation towards a data- and analytics-driven approach to factory operation, as Industry 4.0 implementations become established in every modern factory operation. At its heart is the desire to profit from the exploitation of ‘big data’.

New analytics software has begun to transform the way that industry operates and maintains factory equipment and premises. The insights from analytics are often most profound when they uncover patterns in apparently disparate sets of data.

The more data, and the more types of data, that can be reliably captured from devices in the factory, the more opportunities there will be for software to support advanced functions such as condition monitoring and predictive

28

April 2020

maintenance. The low data bandwidth of the 4-20mA and HART interfaces and the limited scope for integrating them into enterprise computing infrastructures has traditionally hampered efforts to apply analytics to these legacy end-points. It has also restricted the amount of power that can be supplied to an end-point, and the scope to manage the device’s operation remotely.

10BASE-T1L connectivity promises to extend the productivity and efficiency benefits that can be derived from that data to the remotest corners of factories and process plants where sensors and other end-points operate today out of reach of the enterprise network.

The case for installing 10BASE-T1L equipment today rests on the set of capabilities provided for within the 802.3cg standard. A 10BASE-T1L connection offers: • A maximum data rate of 10Mbits/sec

www.controlengeurope.com

over a cable length of up to 1km.

• Up to 500mW of power to endpoints in Zone 0 intrinsically safe applications, enabling the operation of a much wider range of more sophisticated end-points than a 4-20mA or HART system can support.

It can also supply up to 60W of power to non-intrinsically safe applications, depending on the cabling.

• Potential to reuse existing, installed single twisted-pair cabling.

• Rich device management options, including the supply of diagnostic data from the connected device, and the provision of software updates to it.

• An IP (Internet Protocol) address for every node, extending ‘Internet of Things’ capability to the edge of the factory network. An IP address enables a node to be not only monitored but also managed remotely.

Control Engineering Europe

INDUSTRIAL ETHERNET

• Integration with enterprise network infrastructure.

From a hardware standpoint, implementation of 10BASE-T1L equipment is normally straightforward because the physical medium for its communications is single twisted-pair cable. This might even be the same wiring which already carries 4-20mA or HART communications. The 802.3cg standard also supports installation in hazardous (explosion-proof) environments.

It is likely that early implementations of 10BASE-T1L will be of hybrid equipment which supports both the legacy interface, such as 4-20mA, and the new industrial Ethernet protocol.

Making it a success Two critical factors will determine whether a 10BASE-T1L project is successful – a focus on data and network security.

Once engaged in the operational details of a 10BASE-T1L roll-out, engineers can easily lose sight of the reason for it: to lift the veil on the operation of end-points such as sensors, and feed rich streams of data from them to enterprise-level data analytics engines.

It follows that the biggest risk to the success of a 10BASE-T1L project is not at the end-points themselves, or at the physical infrastructure: the problem is most often at the back end, when inadequate provision is made for handling and using the data sets coming from the newly connected end-points.

So, industrial engineers embarking on a 10BASE-T1L installation should keep these questions in mind: • What types of insights do I plan to derive from the data that will be acquired from sensors and other end-points?

• How will the data be integrated into enterprise-level control systems? Is the format of the data from endpoints compatible, or does it need translation?

• How will insights from data Control Engineering Europe

analytics lead to process or system improvements?

The second crucial issue relates to security. The nature of the threat to end-points changes dramatically as soon as they are connected via a 10BASE-T1L network. Before, when connected via 4-20mA, the only way that an end-point could be ‘hacked’ was through physical interference with the device itself or the wires connected to it. A 4-20mA connection is immune to networkborne threats.

The superior connectivity provided by the 802.3cg standard – including an IP address for every node – makes every end-point vulnerable to remote attack via the enterprise network.

The inherent, physical firewall which isolates 4-20mA or HART end-points from the network disappears as soon as the factory installs 10BASE-T1L. This means that individual nodes and the network infrastructure itself have to be secured through the implementation of software technologies such as: • Secure authentication of devices via encrypted device IDs. • Encryption of data transmissions.

• Firewalls to bar outside entities from gaining access to secure devices.

Learning lessons Following the ratification of the 802.3cg standard, the development of 10BASE-T1L-compatible components and equipment has been accelerating.

For its part, Analog Devices (ADI) has been working with industrial equipment manufacturers to ensure that they are able to follow their roadmaps for the introduction of systems that support 10BASE-T1L networking. The expectation in the industry is that products offering 10BASE-T1L capability will be released to the market by mid-2021.

ADI’s experience in supporting customers’ implementations of new technology will help make these 10BASE-T1L product introductions successful. The structure of its Industrial Automation division supports

www.controlengeurope.com

technology implementation marrying technical expertise with market insight to produce the right outcome for the customer. In the case of 10BASE-T1L, this approach will encompass the provision of physical layer products and support for the full communications stack. It also takes account of the long commercial lifetimes of industrial products, backed by a roadmap which forecasts continuous production with compatible 10BASE-T1L products for decades to meet the expectations of industrial customers.

The rapid development of 10BASE-T1L components is enabling industrial equipment manufacturers to start to develop new industrial Ethernet-enabled products. Backed by a consortium of industrial companies which has supported the standard development process, 10BASE-T1L technology looks set to supplant the 4-20mA and HART interfaces and accelerate the adoption of Industry 4.0. plus-circle Brendan O’Dowd is general manager, Industrial Automation at Analog Devices.

More about the IEEE 802.3cg standard The IEEE 802.3cg standard for 10BASE-T1L enables 10 Mbps communication and power up to 1km over a single twisted pair cable.

It is expected that this technology will replace traditional 4 mA to 20 mA or bipolar analogue voltage communications that proliferate within field devices today.

10BASE-T1L provides up to 500 mW of power in intrinsically safe applications and up to 60 W (cable dependent) in non-intrinsically safe applications. The standards provide unified communication and power protocols, with a common networking infrastructure for edge nodes.

April 2020

29

EDGE DEVICES

From the cloud to the edge Dr. Christopher Anhalt explains the difference between cloud and edge computing and how they can work together for the successful implementation of IIoT solutions.

T

he implementation of IIoT applications requires deep integration of production and business processes. Typically it is gateways that connect the automation networks with the IT world, with a central cloud platform acting as IT infrastructure for data collection and data analysis tasks.

However, this architecture has limitations when mapping the specific requirements for individual IIoT applications to the functionality available with cloud platforms.

Firstly, the transfer of huge amounts of production data to the cloud can quickly exceed the capacity provided by the network infrastructure. There is also a requirement to continue processing machine data even if the cloud is temporarily unavailable or the automation process is currently offline. Also, when it comes to critical process data, applications often require a real-time reaction and this is not possible in a scenario where data is first transferred to the cloud to be processed. There will always be some element of latency in such a response.

The solution is to deploy software modules remotely on edge devices, which are also deeply integrated into cloud platforms. Microsoft Azure IoT Edge or Amazon AWS IoT Greengrass, for example, allows certain workloads to be moved to the edge of the network, resulting in a more effective communication with the cloud, a quicker reaction to local data changes and a reliable operation even in extended offline periods. These local IoT edge devices can run selected cloud software modules and access

30

April 2020

IoT edge devices connect the automation and the cloud level, running cloud applications remotely with access to local data.

automation data directly and in (near) real-time via its gateway functionality.

At the same time, the cloud platform is used for tasks such as device management, analytics, and durable storage of data.

Acting as a link From an organisational perspective, edge services and the management of software components at the edge act as a link between the functional requirements of the shopfloor and IT. Roll-out and the entire life cycle of IoT solutions can be managed more efficiently, compared to solutions which require installation, configuration and updates of software locally and individually in each production facility. On the other hand, the decision to use a cloud platform – including edge services – is a strategic one that may not always be easy to take organisational changes may well be required to

www.controlengeurope.com

successfully coordinate the activities at production and IT level.

Softing Industrial Data Intelligence has risen to this challenge – offering software modules for edge computing applications in brownfield installations or data integration in modern plants.

The data acquisition layer provides access to both process and machine data from a range of controllers and devices. The data integration layer uses OPC UA to aggregate data from many data sources and offer additional security. It acts as an abstraction layer, standardising technical differences between applications and unifying their interfaces towards IT. It also can be used as abstraction layer to iron out differences between production sites and offers a unified interface for IoT platforms and other applications. plus-circle Christopher Anhalt is business development manager at Softing Industrial Data Intelligence.

Control Engineering Europe

PRODUCT FORUM •

www.controlengeurope.com to read the full story

Alarm systems management

Complex industrial systems require complex control systems – but carefully thought out alarms systems EEMUA is the acknowledged leader in the field, with EEMUA 191, ‘Alarm systems - a guide to design, management and procurement’, being regarded as the benchmark in alarm systems management.

The EEMUA Alarm Systems e-learning module provides an introduction to EEMUA 191 and is positioned at the awareness level. It offers simple and practical guidance to managers, designers, supervisors and operators on how to recognise and deal with typical human-factor problems involving alarm systems. Its scope covers many sectors, including the energy, process and

utilities industries.

The e-learning is recommended for both discipline and project-focused engineers from a variety of backgrounds who want to gain an introduction to the fundamental principles for design, management and procurement of alarm systems. The course is also relevant to engineers and managers from operating companies as well as specialist contractors and equipment suppliers.

Visit the EEMUA website for further details. www.eemua.org

Creating a successful control environment Know what you want, plan what you’ll get, check that you’ve got it!

The EEMUA Control Rooms e-learning module provides guidance to engineers and the wider teams involved in the design of control rooms, control desks and consoles. It will help during newbuild and modification projects, as well as evaluating existing set ups where people operate industrial processes and activities on facilities such as chemical plants, power stations and oil refineries.

The e-learning will benefit anyone with an interest in process plant control rooms and control desks using Human Machine Interfaces.

It is especially relevant to control engineers, control room console (and HMI) designers and vendors, control room

operators, engineering consultants, engineering contractors, engineering managers, facilities managers, graduate engineers, plant operations managers, process safety managers, SCADA engineers and systems support managers.

Visit the EEMUA website for further details.

The e-learning is positioned at the awareness/introductory level and is an optional precursor to working through EEMUA 201, ‘Control rooms: A guide to their specification, design, commissioning and operation’. www.eemua.org

Assessing Cyber Security for Industrial Control Systems EEMUA’s Cyber Security e-learning course offers flexible online training and certification at the awareness level to benefit all those interested in cyber security in industrial settings.

Leveraging EEMUA’s Cyber security assessment process for industrial control systems, the course covers: The problem • Standards and legal aspects • Malware • Conventional IT versus IACS • Changing challenges • System vulnerability • Human vulnerability • Data vulnerability • Managing machines • Encryption • System countermeasures • Least access principle • Logs review • Managing machines • Cyber security assessment process

EEMUA e-learning for more information please contact us at online-learning@eemua.org Take your place on EEMUA’s Cyber Security Seminar on 25 June 2020 – Register at www.eemua.org Control Engineering Europe

www.controlengeurope.com

April 2020

31

| AT11-14E |

The faster and simpler way to a better machine: with XTS The XTS advantage circulatory movement flexible modular system individually movable movers

User benefits minimised footprint software-based format change improved availability increased output shortened time to market

www.beckhoff.com/xts All over the world product manufacturers have to offer increasingly individualised products – with machines that reduce the footprint and improve productivity at the same time. This is made possible by the XTS eXtended Transport System in combination with PC- and EtherCAT-based control technology. Its high level of design freedom allows new machine concepts for transport, handling and assembly. In the stainless steel hygienic version the XTS is ideal for use in the pharmaceutical and food industries. free installation position compact design freely selectable geometry few mechanical parts and system components

Hall 9, Booth F06