8 minute read
AI in ASW – anti-submarine warfare
– ANTI-SUBMARINE WARFARE
60 Simon Michell talks to Rear Admiral René Tas, Assistant Chief of Staff for Capabilities at NATO Allied Command Transformation (ACT), to find out about the importance of data and artificial intelligence (AI) to anti-submarine warfare (ASW) activities
(PHOTO: NATO/MARCOM)
Q
What opportunities will autonomous systems offer maritime operations?
A Within the maritime domain, one of the most challenging types of warfare is anti-submarine warfare (ASW). The underwater environment has its own challenges, but also offers its own opportunities. We see opportunities in the development of unmanned systems for ASW, for mine countermeasures (MCM) activities and for maritime ISR (intelligence, surveillance and reconnaissance) in general. Not only will these systems collect data, they will require data, and they will process and communicate data.
With this, comes the requirement for machine learning (ML) and artificial intelligence (AI) features in what will become semi-autonomous or fully autonomous systems. Autonomous systems will offer the opportunity to develop and deploy intelligent networks achieving high performance for a wide range of applications in the maritime domain, including ASW. At the same time, we should be aware that such systems will also have operational and technological constraints.
Q
How is ACT supporting the use of AI in ASW?
A Effective autonomy-enabled ASW will require intelligent platforms with state-of-the-art sensors and algorithms that allow optimized detection and tracking capabilities, with reduced human intervention. To this end, ACT is investing in developing advanced assets and algorithms, including AI-based solutions, and testing them at sea. Future capabilities will increasingly use cognitive sonar systems, data fusion, deep convolutional neural networks to self-learn and improve performance as surrounding parameters (such as the environmental ones) evolve. Challenges and opportunities in this field are among the drivers for NATO Allied Command Transformation‘s (ACT’s) Warfare Development Agenda.
Q
How important is data to future ASW capabilities?
A I believe, across NATO, we value the strategic importance of data and we understand that data science is a key enabler to improve our warfighting capability. However, data and data science transcend AI. It starts with data itself. What is the problem to be solved? What data is needed to better understand the problem? How do we obtain it, and how do we share it? We need to answer these questions before we can start with concepts such as machine learning, AI and even algorithmic warfare. It is not just about technology; people, culture, including organizational behaviour, infrastructure and processes, are equally important. And, of course, we need the right policies to make this work.
There is no need to explain the importance, and even the growing importance, of the maritime domain. NATO’s supremacy in the maritime domain is a key and enduring strategic objective. Accomplishing this has become more challenging as, on the one hand, our competitors have improved their capabilities, while on the other hand we have not kept pace with our investments in the maritime domain. Hence, we need to strengthen this investment and, where possible, leverage data, including AI, to enable us to – among other things – outthink, overmatch and outfight our adversaries. It will be no surprise that we need a strong Science and Technology (S&T) foundation to achieve this.
Q
How can AI algorithms help with ASW?
A The operational advantages of employing systems with increased levels of autonomy is that these systems will allow us to act more effectively and efficiently through machine-to-machine and human-tomachine coordination. The AI within systems and platforms with autonomous functions allows an increase of real-time data that can be acquired and that is processed unsupervised at the sensor edge. If we then communicate the result in a secure operational network, the operators – who will be assisted by data fusion features, visualization technologies and automated analysis support – will be faster and smarter than ever before. It will lead to a faster decision cycle, delivering effects more rapidly and effectively. The technology will enable all of this. At the same time, it is important that
Artificial intelligence enables an increase in the acquisition of real-time data, leading to faster decision-making (PHOTO: NATO) 61
NITECH ››› ENSURING ACCESS TO CRITICAL DATA we understand what AI does and can provide in our decision-making. At all times, we must maintain knowledge at all relevant (operational) levels of what tasks are performed by systems, so we build the trust that is required to rely on such systems.
Q To what types of ASW activities are AI or autonomous algorithms being applied?
A Traditionally, ASW has been performed by surface and undersea manned vessels equipped with advanced sensor systems. Autonomous intelligent systems offer the opportunity to develop and deploy an intelligent network that achieves a high-performance capability for a wide range of applications in the maritime domain (airborne, shipborne submarine and sonobuoys), while having less manned systems in harm’s way. Those technologies introduce new possibilities, but also new challenges, in particular for navigation, signal processing, data fusion, underwater communications and networking. These challenges are what ACT and the Centre for Maritime Research and Experimentation (CMRE) are tackling.
Q
How important is teamwork in this realm?
A We require teamwork. That is why it is important that ACT, through its maritime programme of work and its relationship with CMRE, closely cooperates with the multinational NATO ASW Barrier SDI 1.1271 activity and other maritime unmanned system (MUS) initiatives. Equally important is to ensure that we develop capabilities together with the operational community and test new capabilities and concepts together. That is why it is important that all stakeholders are invited to participate in Exercise Dynamic Messenger 2022 where we will test unmanned systems in an operational scenario. War-fighters will be the ones that will use the systems and we, as NATO’s warfare development command, should provide them with the best equipment and processes we can develop.
Q
How is ACT supporting CMRE in this field?
A ACT sponsors CMRE as a contracted science and technology provider supporting an ASW Programme. This includes the technology development and the understanding of the performance space over the different ASW mission areas. We work with CMRE on architecture, networks of autonomous ASW and MCM systems and ISR systems, in order to be able to develop and test capabilities that improve maritime situational awareness and counter threats in the underwater domain. It is about keeping the science and technology edge so we can leverage science and technology results in our warfare development activities, in order to provide the best assets to the war-fighter. The way forward should aim to increase the number of demonstrations and experimentations in both standalone events and operational exercises, to link science and technology results to both directed and open innovation elements, and to further improve cooperation with (multi-)national initiatives.
Note: Admiral Tas will be promoted to Vice Admiral and appointed Commander of the Royal Netherlands Navy starting in September 2021
62
ACT is investing in advanced assets and algorithms, and testing them at sea (PHOTO: NATO/MARCOM)
DEVELOPING ANTI-SUBMARINE WARFARE AI
Robert Been, Principal Scientist and ASW Section Head at NATO’s Centre for Maritime Research and Experimentation (CMRE), highlights the value of collaboration with NATO, multinational initiatives and NATO entities, such as the NCI Agency
As an externally funded organization, NATO’s Centre for Maritime Research and Experimentation (CMRE) has several customers, of which Allied Command Transformation (ACT) is its biggest, and for which it undertakes maritime science and technology work. For example, CMRE’s Dr Alberto Alvarez is working on maritime environmental knowledge and operational effectiveness for ACT. CMRE also develops and tests systems for vessel tracking, mine countermeasures, and, of course, anti-submarine warfare (ASW).
Robert Been, Principal Scientist and ASW Section Head at CMRE, leads a multidisciplinary team working on the development and at-sea testing of advanced autonomous unmanned systems that comprise cognitive sonar components and advanced algorithms to tackle the challenging field of ASW. The team also supports exercises with operational research and dataanalysis expertise. “We will showcase some of this work this September in the operational exercise known as Recognised Environmental Picture, Maritime Unmanned Systems, REP(MUS) 2021, and plan to do so again next year in Dynamic Messenger 2022”.
Data is key in these activities, and CMRE has a vast collection of ASW data in relevant (manned and unmanned) scenarios and areas. This enables the development of machine-learning (ML) algorithms to enable unsupervised processing of sonar data, supporting the end user to take decisions at a much faster pace than in current, mostly conventional, ASW operations. “We apply these algorithms in the domain of target classification, underwater acoustic communication, data fusion and underwater acoustic propagation. The latter two science and technology areas are in close collaboration with Dr Alvarez’s team”.
CLOSE COLLABORATION
In addition, Been’s team collaborates closely with CMRE’s Engineering and IT Division (EITD), which works on scientific networks, software development and data management. In the maritime domain, both the Research Division (RD) and EITD teams, with their in-house expertise and an extensive ASW database, are in a strong position to do machine learning and, thereby, support ACT to achieve their ambitions in that field. The ASW team actively seeks collaboration with the NATO Member States, among others, through the Distributed Autonomous Networked Systems (DANS) initiative, and works with the ASW Smart Defence Initiative (SDI) and Maritime Unmanned Systems Initiative (MUSI). It is also exploring collaboration with the NCI Agency, for ACT, in order to boost machine learning in ASW by exploiting the knowledge base that is present in both CMRE and the NCI Agency.
In spring 2021, Been, together with with RD and EITD colleagues, had a teleconference with the Head of the NCI Agency’s Innovation and Data Science, Dr Michael Street, and NCI Agency Senior Scientist Ivana Ilic Mestric, to see how they could benefit from both organizations’ expertise. One of the options discussed was to run a small pilot project advising ACT on how to take ASW-specific data science and machine learning forward for NATO. 63