[Title will be auto-generated]

Construction Industry Institute®

Leveraging Technology to Improve Construction Productivity

Research Summary 240-1

Construction Industry Institute The AES Corporation Abbott Air Products and Chemicals Alcoa American Transmission Company Amgen Anheuser-Busch Aramco Services Company Archer Daniels Midland Company BP America, Inc. Bristol-Myers Squibb Company CITGO Petroleum Corporation Cargill, Inc. Chevron Codelco-Chile ConocoPhillips DFW International Airport The Dow Chemical Company DuPont Eastman Chemical Company ExxonMobil Corporation General Motors Corporation GlaxoSmithKline Hovensa L.L.C. Intel Corporation International Paper Kaiser Permanente Kraft Foods Eli Lilly and Company Marathon Oil Corporation NASA NOVA Chemicals Corporation Naval Facilities Engineering Command Occidental Petroleum Corporation Ontario Power Generation Petroleo Brasileiro S/A - Petrobras Praxair The Procter & Gamble Company Progress Energy Rohm and Haas Company Sasol Technology Shell Oil Company Smithsonian Institution Solutia Southern Company Sunoco Tennessee Valley Authority Tyson Foods U.S. Architect of the Capitol U.S. Army Corps of Engineers U.S. Department of Commerce/NIST/Building and Fire Research Laboratory U.S. Department of Energy U.S. Department of Health and Human Services U.S. Department of State U.S. General Services Administration U.S. Steel Vale

AMEC AZCO Adolfson & Peterson Construction Aker Solutions Alstom Power Atkins Faithful & Gould Autodesk BIS Frucon Industrial Services Baker Concrete Construction Barton Malow Company Bateman Engineering N.V. Bechtel Group Black & Veatch Bowen Engineering Corporation Burns & McDonnell CB&I CCC Group CDI Engineering Solutions CH2M HILL CSA Group Day & Zimmermann International Dresser-Rand Company Emerson Process Management Fluor Corporation Foster Wheeler USA Corporation GS Engineering & Construction Corporation Grinaker-LTA/E+PC Gross Mechanical Contractors Hargrove and Associates Hatch Hill International Hilti Corporation JMJ Associates Jacobs KBR Kiewit Power Construction McDermott International, Inc. M. A. Mortenson Company Mustang Engineering R. J. Mycka Parsons Pathfinder LLC Pegasus Global Holdings Primavera Systems S&B Engineers and Constructors SNC-Lavalin The Shaw Group Siemens Energy Technip URS Corporation Victaulic Company Walbridge The Weitz Company Worldwater & Solar Technologies WorleyParsons Yates Construction Zachry Zurich

Leveraging Technology to Improve Construction Productivity

Prepared by Construction Industry Institute Leveraging Technology to Improve Construction Productivity Research Team Research Summary 240-1 October 2008

CII would like to thank FIATECH for its generous financial support of this research project. We would also like to acknowledge the efforts of joint CII–FIATECH participants.

© 2008 Construction Industry Institute™.

The University of Texas at Austin.

CII members may reproduce and distribute this work internally in any medium at no cost to internal recipients. CII members are permitted to revise and adapt this work for their internal use provided an informational copy is furnished to CII. Available to non-members by purchase; however, no copies may be made or distributed and no modifications made without prior written permission from CII. Contact CII at http://construction-institute.org/catalog.htm to purchase copies. Volume discounts may be available.

All CII members, current students, and faculty at a college or university are eligible to purchase CII products at member prices. Faculty and students at a college or university may reproduce and distribute this work without modification for educational use.

Printed in the United States of America.

Contents Chapter Executive Summary

Page v

1. Introduction

1

2. How Technology Has Influenced Previous Construction Productivity

5

3. Technology Field Trials

12

4. Technology Prediction Tool

23

5. Prioritized List of Technologies

35

6. Conclusions and Recommendations

37

Appendix

39

Executive Summary Successful use of technology to improve construction productivity involves more than the technical characteristics of the technology. Many factors simultaneously have an impact on construction productivity, including work force characteristics and management practices as well as innovations in technology. To control for such external factors, a threepronged approach was used in this study to develop recommendations, guidelines, and procedures for effectively leveraging technology to improve construction productivity.

First, the study examined how historical changes in construction equipment, materials, and information technologies influenced improvements in construction productivity. Historical data that were analyzed demonstrated significant improvements in labor productivity.

The improvements were primarily due to changes in the functional range of construction equipment, reductions in unit weight of selected construction materials, and the automation and integration of project information systems. Labor productivity improvements associated with the use of these technologies ranged from 30–45 percent.

Second, a field test of materials tracking and locating technologies was conducted to measure how using such technologies can improve productivity on prototypical CII construction projects. Radio-frequency identification (RFID) tags and a Global Positioning Satellite (GPS) system were used to track materials in lay down areas for two CII member projects, one in the U.S. and the other in Canada. The field test showed that not only did improvements in material tracking improve productivity at the construction workface and the retrieval of materials in the lay down areas, but also improved the predictability and reliability that materials would be available when needed.

Third, a four-stage predictive model that estimates the potential for a technology to have a positive impact on construction productivity was developed and validated. The model combines results from historical analyses and the field tests into a tool that can be used as a predictor of how future technologies might perform. The model was used to develop v

performance scores based on a historical perspective of 28 technologies.

The technologies were selected to represent cases where construction jobsite implementation was: (1) successful, (2) inconclusive, or (3) unsuccessful. Statistical analysis confirmed that average performance scores produced by the model were significantly different across these three categories. The model was then used to develop a list of promising emerging technologies.

In summary, this research project has demonstrated with significant statistical analysis, field trials, and tool development how technology can be used to leverage productivity.

vi

1 Introduction Advances in technology have many benefits. Among the most often cited are improved quality and productivity. The evolution within the communications industry can be seen as one example of how an industry embraces and leverages innovation on a continuing basis. Arguably, the construction industry lags in this regard and underutilizes advances in technology. The opportunity to improve construction productivity exists, however, and evidence suggests that sectors of the construction industry have experienced long-term productivity growth as a result.

Economic research has shown that technology tends to have a greater impact on labor productivity versus factor productivity measures. For example, investment in new equipment technology may improve an organization’s labor productivity, but its factor productivity may actually decline if the relative increase in the cost of the equipment outweighs the relative savings in labor costs and gains in output. While evidence shows that improvements in equipment and material technology have a positive impact on construction productivity, the impact of information technology (IT) has remained largely undocumented.

IT clearly has the potential to improve construction productivity through improved communication, logistics, planning, and resource allocation, but exactly how it improves construction productivity is not well understood. CII efforts in the fully integrated and automated project processes (FIAPP), the FIATECH consortium efforts as well as a major study by the National Institute of Standards and Technology (NIST) indicate that much can be gained via interoperability alone.

There tends to be a required critical mass in terms of adoption before a new technology has an impact on production. Malcolm Gladwell’s The Tipping Point posits that it takes a series of factors, people, and events to contribute to widespread adoption of an idea, trend, or technology. The adoption of electricity is an example that follows Gladwell’s thinking; while electricity was first introduced in the 1880s, it was not until around

1

1920 when the majority of U.S. industrial machinery became electrically powered. In other words, it is not just a discovery or the advancement of a technology that ensures acceptance and adoption, but rather a series of interrelated events that contribute to the success or demise of an invention. Similarly, it is arguable that the North American construction industry has not reached the critical threshold with regard to widespread implementation and integration of information technology. Certainly, innovations improve the impact of any new technology because they decrease the investment and maintenance costs along with a gain in understanding of how the technology can most effectively be utilized to improve productivity.

Problem Statement Significant value can be gained for construction if technology could be more effectively leveraged to improve productivity. Changes in technology have had a profound effect in other industries and have led to strategies to further promote productivity by leveraging technology usage. Understanding how past technologies have improved construction productivity and demonstrating how new technologies can do the same will help CII member organizations identify emerging technologies that could improve productivity.

Objectives The primary objective of this research was to identify how technology has and can improve construction productivity. Since a multitude of factors simultaneously impact a project’s productivity including work force characteristics and management practices, it is difficult to isolate technology’s impact on productivity. Therefore, the research utilized an activity-based analysis in order to control for the number of factors that influence productivity at the project level. Correlating changes in technology with changes in productivity through statistical analysis allowed the research team to identify the order of magnitude that different technological characteristics have had on construction productivity.

This formed the foundation for a model to assist future planning and implementation of new technologies. The model was validated using

2

different approaches and was expanded through field testing of a material tracking system, which occurred at the same time as the model was developed. By conducting the field tests, this project allowed CII to achieve multiple objectives and to continue a thread of research that is needed among CII member organizations. Objectives and corresponding deliverables include: 1. Characterize and quantify the impact (in orders of magnitude) that changes in equipment, material, and information technologies have had or can have on construction productivity; 2. Develop a model that utilizes the research analyses to predict the potential impact of current and new construction technologies on productivity; and 3. Execute field testing of a technology (in this case, automated materials tracking and locating) to help validate the predictive model by incorporating characteristics of information technology’s impact on construction productivity.

Research Methodology A considerable amount of knowledge exists on the subject of technology and productivity. The research team conducted a thorough literature and background review, including use of extensive information from CII’s knowledge structure along with previous research from the Center for Construction Industry Studies (CCIS), FIATECH, and academic journals. The review focused on: 1. Changes in equipment, material, and information technologies that had an impact on construction productivity.

2. Technology drivers and their impact on productivity improvement.

3. Advances of relevant technologies in manufacturing and other industry sectors, which may have application and/or provide productivity benefit in the construction industry.

In a previous study, the research team’s principal investigators collected longitudinal productivity data spanning three decades on 300 construction activities. This data, combined with data from the CII

3

Benchmarking and Metrics Productivity Database, were used in this study to examine productivity improvements over the past two decades.

With focus on selected activities, technological advances in materials, equipment, and information technologies were identified and the respective impacts on productivity were assessed.

Historical productivity data and expert opinions were used to develop a predictive model. The model is intended for CII members aiming to assess the suitability of emerging technologies for their respective organizations. The resulting model was designed to assist in the evaluation of the viability, challenges, and benefits of new technologies.

The predictive model was validated using different approaches.

To further validate the predictive model and directly observe how a technology can improve construction productivity, the research team concurrently conducted field trials of an emerging material tracking technology. This task included the development of the field trial methodology and the selection of candidate sites. For this effort, the research team focused on the use of radio-frequency identification (RFID) tags and global positioning systems (GPS) to track engineered components in construction job sites. The predictive model was also used to prioritize a list of emerging technologies that have the highest potential to improve construction productivity.

4

2 How Technology Has Influenced Previous Construction Productivity The research examined technology’s influence on construction productivity in two parts. First, the relation between construction equipment and material technologies and construction productivity were examined. This was followed by examining the influence of the automation and integration of information technology on construction productivity. Together, understanding these historical trends helped the team understand why current technologies are improving jobsite productivity and to predict which future technologies can likely do the same.

Equipment and Material Technology Construction equipment and materials have seen significant changes over the past several decades. However, the construction industry has lacked documented evidence of how improved technologies in construction have resulted in productivity improvement. This relation was examined by measuring productivity changes over time for individual construction activities and then comparing equipment and material technology changes for these same activities. For example, concrete placement productivity measured in cubic yards per hour was compared from rates achieved in 1977 to placement rates achieved in 2004 to arrive at an overall percentage of change in labor productivity.

A similar comparison was made for 300 other construction activities, after which comparisons were made between productivity rates and technology changes in the activities. The primary source for productivity rates was the Means Building Construction Data, Richardson Process Plant Estimating Guides, and Dodge Unit Cost Books, which are common industry references used for estimating the cost of construction projects.

To examine equipment technology’s influence on construction labor productivity, a longitudinal study examined the changes in equipment technology and labor productivity of 200 construction activities over a

5

22-year period (1976-1998). Four factors were identified (defined below and examples discussed later) that characterize significant changes in equipment technology related to improvement in the activities’ labor productivity: • Amplification of Human Energy – involves technology designed to make an activity physically easier to perform. In its simplest terms, it can be regarded as the shift in energy requirements from human to machine and causing an increase in machine output (e.g., RPM, HP). As an example, welding machines increased wattage output and powder-actuated systems offered greater depth penetration for installing studs in metal decking. In addition, cranes developed greater lifting capacity over the 22-year period. Further, most site work machinery offered increased horsepower output (e.g., frontend loaders, dump trucks, backhoes, bulldozers, graders, asphalt pavers, and scrapers).

• Level of Control – relates to advances in machinery and hand tools that transfer control from the human to the machine.

Welding machines in the metals division, for instance, are now equipped with remote controlled amperage adjusters and powder actuated systems have semi-automatic loading capabilities. The pneumatic nail gun has replaced the handheld hammer in the wood and plastic division as well as in formwork installation in the concrete division. Also in the concrete division, pump trucks are now equipped with remote-controlled booms, and concrete vibrators automatically adjust the vibration frequency to match the concrete’s slump.

• Functional Range – expands the range of capabilities of a tool or machine. Through advances in hydraulic controls and microprocessors, machinery for sitework now offers more control precision and a greater reach with booms and buckets.

Excavators and backhoes are also capable of digging deeper.

• Information Processing – advances here have occurred in heavy machinery with the development and improvement in engine performance monitoring and self-diagnosis systems.

6

Previous research on 100 additional activities over a 27-year period (1977–2004) focused on examining the relation between changes in construction materials and labor productivity. Three primary factors were considered as effective material characteristics: 1. Modularization – relates to the amount of material customization performed onsite prior to installation. Pre-fab of individual components was also considered in the category of modularity.

The purpose of including this factor is to measure the benefits of offsite construction that allow more efficient installation when materials can be “customized” in an environment (offsite) under ideal conditions before actual installation.

2. Reduction in unit weight – includes ease of handling and transporting by craft labor, although lighter materials have other benefits related to structural design and space requirements.

3. Installation flexibility – refers to the environmental conditions under which a material can be installed. For example, extreme temperature or moisture can have significant impacts on installation. Technological advancements have improved the durability of materials and allowed installation in extreme moist and cold conditions. The installation improvements were attributed to technological advancements in materials such as epoxy coating, water-proofing, and cold weather admixtures.

The activities that experienced improvement in the above equipment and material technology traits experienced more improvements in labor productivity than those activities that did not, and this finding was statistically significant (Table 1). Activities experiencing an improvement in energy, control, functional range, and information processing had at least twice as great of an improvement in labor productivity than activities experiencing no improvement in the technology factors. Likewise, it was found that activities experiencing improvement to modularization, unit weight, and installation encountered at least twice the improvement in labor productivity compared to those activities with no likewise material improvements.

7

Table 1. Change in Equipment and Material Technology versus Changes in Labor Productivity Technology Characteristic

Change in Labor Productivity No Change in Equipment Technology Characteristic

Change in Equipment Technology Characteristic

∆

Energy

3.6%

39.8%

36.2%

Control

14.9%

46.6%

31.7%

Functional Range

13.5%

51.8%

38.3% 35.4%

Equipment Technology Characteristic

Information Processing

21.0%

56.4%

Material Technology Characteristic

No Change in Material Technology Characteristic

Change in Material Technology Characteristic

∆

Modularization

8.1%

24.2%

16.1%

Reduction in Unit Weight

10.4%

48.6%

38.2%

Installation Flexibility

8.7%

23.1%

14.4%

Automation and Integration of Information Systems The automation and integration of information systems are built around predefined tasks/work functions common to most projects. In this context, automation technologies focus on the degree to which data of individual work functions (such as supply management and project management) are automatically manipulated for the use by other work functions. Integration technologies focus on the ability to exchange information between work functions and their associated databases (e.g., exchanges of information among supply management and project management functions).

Data from the CII Benchmarking and Metrics (BM&M) Productivity Database were used by the research team to measure the impacts of information systems’ automation and integration technologies on jobsite productivity. The BM&M data are collected on a project level, where each project includes information on raw productivity metrics, cross-

8

referenced to technology usage and CII work functions. The research examined the relationship between jobsite productivity across four trades (concrete, structural steel, electrical, and piping) and the automation and integration of various CII work functions. It is important to note that based on CII’s productivity definition a lower productivity number is better.

Furthermore to ensure organization confidentiality, the raw productivity metrics were normalized between 1 and 10 using the max–min method of normalization.

A project-level analysis was performed using two composite indices that summed automation and integration levels across these CII work functions: • business planning and analysis • conceptual definition and design • project definition and facility design • supply management • project management • offsite/pre-construction • construction • as-built documentation • facility start-up and life cycle support Project management, due to its importance to the project execution process, is further subdivided into five work functions: coordination system; communications system; cost system; schedule system; and quality system. The relationship between normalized productivity for each work function and automation/integration technologies was then analyzed for each trade-specific sample dataset. For purposes of the analysis, projects scoring 5 percent above the overall median were classified as having a high level of automation or integration, and projects scoring 5 percent below the median were defined as having a low level of automation or integration.

9

The results of the analyses involving the automation and integration of the CII work functions and their correlation with project productivity are shown in Table 2. When work function and productivity data from all trades were combined, increased automation and integration of the CII work functions was strongly associated with better productivity performance, and the difference was statistically significant. As a reminder in this case, a lower productivity measure is better. Furthermore, the study examined the productivity difference by the concrete, structural steel, electrical, and piping trades by whether projects’ high versus low level automation and integration of the CII work function, and the analyses observed that better productivity was more frequently observed on projects with higher automated and integrated work functions.

Table 2. Automation and Integration of CII Work Functions versus Labor Productivity Normalized Labor Productivity High Level Automation

Low Level Automation

∆

All Trades

3.7

4.5

-0.8

3.5

4.8

-1.3

Concrete

3.5

3.9

-0.4

3.0

4.5

-1.5

Structural Steel

3.7

5.2

-1.5

3.7

4.3

-0.6

Electrical

3.6

5.2

-1.6

3.5

6.4

-2.9

Piping

4.0

4.4

-0.4

3.9

4.7

-0.8

Trade

High Level Low Level Integration Integration

∆

The normalized productivity measures preserve the confidentiality of the CII BM&M data and also allow comparisons across different tasks and trades, since the measures become dimensionless. To help clarify the results, the means of raw productivity for each trade-specific sample dataset were calculated for the projects with high and low level use of automation and integration information technology (Table 3). The percentage difference between the means was calculated. The percentage difference of productivity indicates the percentage of time savings per installed quantity when using a high versus a low level of automation and integration. The same procedure was used for each trade-specific

10

sample dataset to analyze the differences between a high versus a low level of integration. While both the integration and automation of the CII work functions were related with better productivity performance, the analyses suggest that integration has a greater impact.

Table 3. Percentage Improvement in Raw Labor Productivity Measurements Considering Automation and Integration of CII Work Functions Percent Improvement in Labor Productivity Automation

Integration

30.9%

45.0%

Concrete

23.3%

56.4%

Structural Steel

33.9%

41.5%

Electrical

30.3%

38.4%

Piping

36.4%

45.9%

All Trades

While skills and technology appear to positively affect labor productivity, the declining skill level of the construction work force may become an obstacle to, but motivation for, the widespread implementation of future technological advances in the U.S. construction industry. Future technological advances in construction will need to consider the availability of skilled workers to use advanced construction machinery, for example. At the same time, the declining skill level of construction workers may create the demand for more technological advances.

Therefore, the construction industry should address the development and implementation of advanced technologies and also consider the skill level of the construction work force.

11

3 Technology Field Trials Examining archived records of past productivity and corresponding usage of technology helps explain how technology can improve construction productivity. Directly observing how a new technology improves productivity provides rich detail regarding the implementation, craft worker and management perceptions towards new technology, and other intangible factors that significantly influence a technology’s impact. This study’s direct observations involved a series of field trial efforts involving an emerging materials tracking system used on two projects. One project was located in Rockdale, Texas, and the other in the Portlands industrial yards in Toronto, Ontario, Canada.

The objectives for the field trials were: • Assess the impact of a material identification and locating technology on labor and construction productivity.

• Help to assess the validity of the predictive model developed in this research study.

• Capture lessons learned regarding technology implementation issues and use during the trials.

Justification for Focus on Material Tracking There is significant value to be gained for the construction industry if technology could be more effectively leveraged to improve construction productivity. CII Research Team (RT) 173, Update Construction Technology Needs, identified field materials management as a work process with significant potential to be positively impacted by new technology. CII RT 215, Work Force View of Construction Productivity, recently identified field materials management as one of the most significant factors currently impacting construction productivity among CII member organizations. One RT 215 recommendation was for future research to examine how to improve onsite material availability. By conducting the field tests described in this document, multiple objectives have been achieved and a CII thread of research continues.

12

Technology Overview In the materials identification and localization approach used for these field trials, active RFID tags were attached to selected construction components. As each tag had a unique identification code, tags were uniquely related to the corresponding components to which they were attached. RFID and GPS readers were connected to a handheld computer that was used to collect data on the site. During data collection, an individual carrying this handheld computer drove and walked around the job site. As he or she moved around the site, the GPS receiver determined its own position while the RFID reader identified the presence of the tagged components around that position. Then localization algorithms processed the collected data to estimate the location more accurately for each tagged component. Maps displaying these locations were printed and given to workers so they could more easily search for tagged components. Alternatively, real-time navigation functions could have been used in these trials. Using this navigation mode, the handheld computers would have shown both the position of the worker equipped with the RFID-GPS receiver and the estimated location of the tagged components.

Rockdale Field Trial This field trial was performed at the Sandow Steam Electric Station Unit 5 project in Rockdale, Texas. Bechtel is the contractor and Luminant Energy (formerly TXU) is the owner. The power plant project is a 565 megawatt circulating fluidized bed, lignite-fired power plant, which incorporates state-of-the-art emissions control technologies and consists of two boilers, two bag houses, one stack, and one turbine. The project has two almost identical steel structures (named A and B in this section) to support the steam generation processes. Both structures were composed of approximately 4,800 steel components and were divided in similar sequences of installation. Each boiler structure (Figure 1) had its own assigned cranes, equipment, foreman, and installation crews. The field trial was conducted from August 1 to October 19, 2007.

13

Figure 1. Boiler Structure The job site was divided into two main areas for the purpose of this study: the lay down yard and the installation area. The lay down yard stored the structural steel components in an area of 25 acres, while the installation area held the components retrieved from the lay down yard before their installation. The installation area was small and crowded with materials, equipment, and workers. The distance between the installation area and the lay down yard was approximately 1,000 ft.

Original Field Materials Identification and Localization Process Upon receiving, structural steel components were checked against the packing list of the corresponding delivery and unloaded in the lay down yard. The 25 acres of the lay down yard area were subdivided in grids of 50x100 ft; each grid was identified with an alphanumeric code and was referenced with a grid post at its center in order to facilitate the location of the stored components. The corresponding grid where each component laid was manually recorded in the field and, later on, entered into an electronic inventory system. In addition, components were hand-marked

14

at delivery –including the piece marks (i.e., codes) of the components, the delivery date, and a truck identification code. During their storage at the lay down yard, a given component could be moved from one grid to another several times before retrieval for installation. (Traditionally, workers record these movements as they happen to keep an updated record of component locations.) When the components needed for installation were identified, a materials withdrawal request (MWR) list containing these components was submitted by installation foremen.

Then workers had to locate and flag these items based on their grid records and written piece marks. Once flagged, craft workers loaded the components onto trucks and hauled them to the installation area.

At the installation area, structural steel components were unloaded upon arrival in a pre-defined area close to the installation work. Neither the presence (identification) nor the location of the components in the staging area was controlled by the steel erection crews. Instead, they commonly relied on what that they could recall about their positions.

Eventually, the components were installed.

Automated Field Materials Identification and Localization Process Selected structural steel components were tagged at the lay down yard after being received. There, GPS and RFID data were collected on a daily basis in order to generate and update the position of the tagged components. Whenever a new MWR was submitted, a map showing the location of the corresponding components was generated and handed to the lay down yard workers (Figure 2). Then workers proceeded to flag the components illustrated in the maps; later on, craft workers loaded the components onto trucks and hauled them to the installation area.

There, data collection took place once or twice a day, depending on the amount of new received components. For the tagged components present in the installation area, a list with their piece marks and a map with their locations were handed to the foreman in charge, so he/she could more reliably plan for their installation. Before the components were lifted by cranes, the erection crews removed the tags from the components.

15

Figure 2. Sample Map at the lay down yard Field Trial Results Boiler A followed the original materials identification and localization process, while a section of Boiler B followed the automated materials identification and localization process. For both boilers, 400 components from similar installation sequences were tracked during the trials and their respective productivity records were collected, leading to the following results: • The average time that workers spent locating the structural steel components on the lay down yard for Boiler A was 36.8 minutes, while the average time to locate the components on the lay down yard for Boiler B components was 4.6 minutes.

This difference in labor times was statistically significant.

• The number of components not immediately found in the lay down yard was reduced by a ratio of 18 to 1 when using the automated process.

• Nineteen percent of the tagged components were moved to a different location in the lay down yard at least one time during the 10-week trial. This reinforced the perception that

automated materials tracking could improve craft productivity and minimize the number of components not immediately found.

• In the installation area, the productivity rate associated with steel erection tasks was increased by 4.2 percent when using the automated process. This productivity data is based on the work hours required to unload, store, identify, and erect steel components on the installation area. It does not include the effort needed to plumb, align, paint, and inspect these components because these activities remained unaffected by the way components were tracked.

Overall, these differences in productivity indicated that the automated process potentially improved the craft productivity of the activities involved.

Finally, the feedback obtained from managers and workers during the trial was extremely positive. According to the materials manager, automated materials tracking is a technology that could potentially save a significant amount of labor work-hours for locating and inventoryupdating purposes. The general foreman’s opinion was that knowing which components are available for installation and their respective locations could be extremely helpful for planning purposes and allow for a more intensive focus on installation rather than on materials availability issues. Site workers repeatedly suggested tagging all the components in the lay down yard so that they could rapidly and confidently locate them and thus avoid site-wide type of searches. Indeed, as the trial went on, both workers and foremen proactively requested the researchers to generate additional location maps for a number of items, especially those that needed to be immediately installed yet could not be found.

Portlands Field Trial The Portlands Energy Centre (PEC) project was a partnership between Ontario Power Generation (OPG) and Transcanada Energy to build a 550 megawatt, natural gas fired, combined cycle, power generation

17

facility (Figure 3). Because of its location on the Toronto waterfront, a port facility was located about one mile away from the site, and roll-on capabilities existed for the delivery of large modules (Figure 4). Initial field trials focused on tracking over 200 pipe spools and safety valves.

Figure 3. Progress Photo as of November 2007 (Photo Source: PEC project website)

Figure 4. Portlands Energy Center (PEC) Layout 18

Original Field Materials Identification and Localization Process The original field materials management process was well-defined by the contractors. Warehouse personnel were responsible for receiving, storing, tracking, and releasing requested materials to subcontractors. A work packaging and expediting group worked closely with the warehouse personnel. Several storage areas were used including a nearby port area warehouse, lay down yards, and staging areas.

Automated Field Materials Identification and Localization Process The original materials tracking processes were augmented for this small-scale field trial. The automated materials tracking technology was used on two subsets of critical components that had caused crew delays on past projects and had negatively impacted project schedules in the past. The trial was implemented with vendor support, the university researchers, and with the assistance of the project personnel. Maps were produced on request that indicated spool and valve locations overlaid on satellite imagery with a translucent project plan view layer used as well for orientation (Figures 5 and 6).

Figure 5. Pipe spools at port

19

Figure 6. Pipe spools at site laydown areas Two subcontractors used the proposed technology in the late summer and fall. Rather than use maps for day-to-day planning, the subcontractors chose to use the maps opportunistically. Some initial scepticism existed.

Maps were requested for selected items after the subcontractors had spent significant crew time searching for the items. In all cases the materials were located immediately for the subcontractors.

Field Trial Results Results from the Portlands field trials addressed the impact of the location and identification technology on labor cost through a series of case studies. During the trials, several instances occurred where pipe spools were requested to be located using the identification and localization technology. The cost associated with locating the technology was quantified with and without the use of the identification and localization technology. A summary of the results from the case studies during the trial follows: • Overall feedback from contractor A was positive after the first successful location and retrieval.

20

• Contractor A immediately recognized benefits in terms of time and cost savings.

• Each location and retrieval was estimated to have saved the project $4000–$5000.

• Contractor A was able to reduce initial crew size from 18 to 12 workers knowing that he would not have to dedicate resources to locating materials or be faced with peak demands in order to make milestones, and the resulting labor savings to the contractor were significant.

• Overall feedback from Contractor A was strong and positive regarding the level of information provided.

• Contractor B would only request locations as a last resort and not as the first step.

In addition to the case studies, initial attempts at using the materials tracking technologies for input into decision support systems were also tried such as in the example in Figure 7.

Figure 7. Number of RFID tags (pipe spools) in each laydown area by week

21

Lessons Learned from the Field Trials Both the Rockdale and Portlands field trials captured valuable lessons learned associated with the deployment and implementation of an identification and localization technology for materials tracking purposes. These lessons learned are summarized here as three main ideas or concepts.

First, project data and sensing data need to be properly integrated.

Data integration from project management systems and the technology results (identification and position information) is needed when trying to control large number of components simultaneously in an almost realtime fashion.

Second, to increase the visibility throughout the supply chain, the RFID tags could be attached to the components at the manufacturing facilities. This practice could allow contractors to effectively know which components are arriving at the site and have an automated notice of this event as it happens.

Finally, a deep reengineering of original field materials management processes would be needed in order to take full advantage of the automated tracking technology. Reengineering a fully integrated and automated process (including the supply chain) would enable potential elimination of intermediate tasks and simplify the management process, creating the highest possible benefit to cost ratio. Estimating the eventual benefit to cost ratio for the automated materials tracking technology presented here is difficult; however, several industry members have already decided to adopt and implement this technology as soon as possible.

22

4 Technology Prediction Tool The predictive model is based on the assumption that past performance can predict future success. Using primarily historical data, the predictive model estimates the potential for a technology to have a positive impact on construction productivity. The model incorporates a series of previous studies as well as analysis of CII Benchmarking and Metrics data by RT 240 to identify technology characteristics that have been shown to have a positive influence on productivity.

Predictive Model Framework The model consists of four stages: I. Strategic Economic Analysis II. Technical Feasibility III. Technology Usage Issues IV. Technology Impact The Strategic Economic Analysis stage explores the technology’s costs, potential benefits, and documentation of the technology being analyzed. Next, the Technical Feasibility stage examines criteria that relate to the technology’s maturity, compatibility with current technologies, and the risks associated with its implementation due to issues such as the technology vendor’s standing in the industry and existing technological standards. The Technology Usage Issues stage examines factors surrounding the likelihood that the technology will be extensively used. Finally, the Technology Impact stage examines specific technology characteristics that have been found to influence construction productivity directly, and accounts for whether the technology is a material, equipment, and/or information technology.

Each stage is further divided into three or more categories with each category broken down to a number of different elements that are phrased as a series of questions that describe the technology being evaluated.

23

Overall, RT 240 identified 79 potentially different elements (i.e., questions), and the team was able to provide historical data to support 55 of the elements (70 percent). Another 15 percent of the criteria, such as Technology Maturity, are cited multiple times throughout technology related literature but do not have supporting historical data. The remaining 15 percent can be found in Stage I, which contains criteria such as: potential benefits, technology costs, and technology documentation.

Although these criterions lack historical data analysis, these factors have merit when considering the implementation of a new technology within the construction sector. A complete list of the 4 stages, 12 categories, and 79 elements are included in Tables 4, 5, 6, and 7.

Table 4. Technology Prediction Tool – Stage I Stage I: Strategic Economic Analysis A. Budget Analysis A1. Capital costs are within the project’s budget.

A2. Implementation costs are within the project’s budget A3. The indirect and overhead costs associated with the technology are within the project’s budget.

A4. An economic analysis is required and has been adequately performed.

B. Potential Benefits B1. Potential for reducing project costs B2. Potential for improving safety B3. Potential for improving project quality B4. Potential for improving quality of information B5. Potential for reducing project duration C. Technology Documentation C1. Documented case study of technology’s benefit exists C2. Positive, statistical evidence of technology’s benefit exists

24

Table 5. Technology Prediction Tool – Stage II Stage II: Technology Feasibility D. Technology Maturity Level D1. Basic principles observed D2. Process with basic package exists D3. Field demonstration completed D4. Functioning versions exist D5. Widely used throughout industry E. Technology Risks E1. Vendor has developed successful past technologies E2. Vendor reputation E3. Dominant vendor exists E4. Safety risks with implementation E5. Organization internal risks E6. Open standards for technology F. Technology Compatibility & Integration F1. Compatibility with current information technology or processes F2. Capable of integrating with future systems F3. Supporting technologies critical for implementation F4. Commercially available supporting technologies F5. Industry implementation plan for technology.

25

Table 6. Technology Prediction Tool – Stage III Stage III – Technology Usage Issues G. Technology Acceptance G1. Perceived ease of use G2. Perceived usefulness G3. Incentive for using technology H. Technology Synergy & Protocol H1. Goodness of fit between the project’s and engineering’s needs H2. Goodness of fit between the project’s and the workforce’s needs H3. Documentation of current work processes.

I. Technology Logistics Functions I1. Use in improving peak workforce availability I2. Use in schedule controls I3. Use in material management I4. Use in contractor’s organization I5. Use in information flow management

26

Table 7. Technology Prediction Tool – Stage IV Stage IV – Technology Impact J. Construction Equipment Technology J1. Level of control

J3. Information processing

a. Manual hand tool

a. No level of information feedback

b. Manually controlled device

b. Feedback on internal operations

c. Remote controlled device

c. Feedback on environmental parameters

d. Computer assisted device

d. On-board computing systems provide supervisory control

e. Autonomous device J2. Amplification of human energy

J4. Functional range

a. No driving energy supplied

a. No enhancement to work envelope

b. Some driving energy supplied

b. Extension of physical range

c. All driving energy supplied

c. Precision enhancement

K. Material Technology K1. Modularity

K3. Unit Weight Reduction

a. 100% on-site customization

a. No change in unit weight

b. Less than 100% on-site customization

b. Unit weight reduced by 0 to 30%

c. No on-site customization

c. Unit weight reduced by 30 to 60%

K2. Installation

d. Unit weight reduced by more than 60%

a. Controlled environmental conditions required b. Less than ideal environmental conditions allowed c. Any expected environmental condition acceptable.

27

L. Integration Potential of Work Functions L1. Conceptual definition & design L2. Project definition & facility design L3. Supply management L4. Coordination system L5. Communication system L6. Cost system L7. Schedule system L8. Quality system L9. Offsite/pre-construction M. Automation Potential of Work Functions M1. Project definition & facility design M2. Supply management M3. Coordination system M4. Communications system M5. Cost system M6. Schedule system M7. Quality system M8. Offsite/pre-construction M9. As-built documentation M10. Facility start-up & life cycle support.

Prediction Tool’s Report and Scoring Mechanisms The prediction tool generates a score as the user answers the questions associated with each of the model’s elements. Each question is answered by using either a Likert scale or a “Yes, No, or Maybe” answer.

The Likert scale utilizes descriptions such as: strongly agree, agree, undecided, disagree, and strongly disagree to describe a characteristic of a technology or the related implementation efforts surround the use of the technology.

As mentioned, most of the elements are based on historical data. The analyses presented in Chapter 2, for example, were used to develop weights for elements in Stage IV. Through literature review, RT 240 identified statistical analyses developed by other researchers and those

28

studies archived in peer-reviewed technical journals and used these analyses to develop weights for elements in Stages II and III. For other elements, the Analytical Hierarch Process (AHP) was used to develop appropriate weights. AHP uses pairwise comparisons based on expert input to develop weights regarding the relative importance of different elements as well as the sections and stages in the model’s scoring mechanism. RT 240 developed default AHP weights, which were used to initially calibrate the model. Individual users, however, are able to adjust the AHP weights based on their own expertise, and as a check the tool measures the consistence in logic if the weights are adjusted.

Ultimately, the model’s score is summarized in the Prediction Report (Figure 8), which graphically portrays the technology’s score for each of the four stages. Below the graphs, the stage’s score, maximum possible score, and stage percentage is displayed. The user is prompted to complete each stage and is either given instructions to move on to the next stage or a warning will appear to let the user know that the technology is out of the predictive tool’s scope if 20 percent of the questions are answered as “Undecided/NA.” The total technology prediction is based

Figure 8. Prediction Report

29

on a percentage, which is displayed in the lower right corner of the prediction report. This percentage is a calculation based on user input and the maximum possible scores (considering the AHP weights) of each individual stage. Just below the total technology prediction is a reliability percentage, which indicates the accuracy of the prediction based on the percentage of elements that are answered by the user.

Tool Validation Using primarily historical data, the predictive model estimates the potential for a technology to have a positive impact on construction productivity. To perform this validation, RT 240 utilized two approaches to analyze the model’s scores.

Validation Approach I Validation Approach I was based on the experience of team members and their peers within their respective organizations. The members of RT 240 used the model to evaluate a technology that was adopted for a certain project or projects. The evaluator only used the knowledge of the technology at the time of pre-implementation. Once a score was recorded within the prediction model, the evaluator assessed the level of success of the technology as either a success, middle of the road (neither a clear success nor a failure), or failure.

Validation Approach II The second validation approach used a hypothetical analysis of technologies that have already been implemented in the construction industry. Recognizing that a common context would need to be used for assessing the potential technology’s implementation, each technology was evaluated based on its potential use on the CII Model Plant at the time when the technology was first developed. Originally developed by CII in 1985, the CII Model Plant has served as a standardized physical facility for productivity measurement. The Model Plant updated by RT 231, Craft Training in the U.S. and Canada (2005-2006). The estimated average cost of the plant in 2006 dollars is $143.6 million. The estimated duration of construction is set at 78 weeks. In order to set thresholds for 30

successful, middle of the road, and unsuccessful technologies, the team identified whether technologies fell within each outcome based on the expertise of RT 240 members.

Once again, a combination of successful, middle of the road (neither a clear success nor a failure) and unsuccessful past technologies was used in the validation. Examples of successful technologies included the obvious such as joystick controls on heavy machinery, global positioning systems, and computer aided drafting.

Combined results of the model’s validation using approaches I and II are shown in Figure 9.

80%

R

Total Percentage Score

R R R R R 60%

R R R R R R R R

40%

R R

Successful Threshold

R R R

R

R R

Emerging Threshold

R R R R

20%

R Successful

Middle of the Road

Unsuccessful

N = 15

N=5

N=8

Figure 9. Box Plot of Validation Results for Approach I & II Lessons Learned from the Predictive Tool and Validations The boxplot diagram, as shown in Figure 9, maps the number of outcomes against the evaluation score. From the diagram, it is apparent that the higher the score, the higher the chance of success. Statistical analysis confirmed that average total percentage scores on the model 31

were significantly different across the categories of success, middle of the road, and unsuccessful. Simplistically, this validates the co-relation between the model’s score and the technologies’ success rate. The thresholds shown on the diagram provide simplistic categorization.

Clear patterns emerged when the combined validation results from Approach I and II were examined. In Figure 9, 90 percent of the technologies that scored above a 60 percent on the model's total score were considered successful. Approximately 50 percent of the technologies that scored between 40-60 percent were a considered a success. Finally, no successful technologies scored below a 40 percent.

Based on these outcomes, RT 240 felt that 40 percent and 60 percent be used as cut-off points for the successful and emerging technology thresholds. Technologies that score between 40-60 percent have potential for being successful and may be on the cutting-edge, but have no guarantee of being fully accepted by industry. Technologies that score above a 60 percent are typically considered successful technologies.

These technologies have been proven and may be in wide use throughout industry.

Based on the criteria used for the predictive tool, it is apparent that as a particular technology matures over time, the evaluation will yield higher scores over time. The question is at what level of maturity does it make sense for an organization to make significant investments? If it starts too early, risk of failure is high, which typically involves lost investment. At the same time, if it waits too long (thresholds over 70 percent), there is a business risk that competitors will pass them by, resulting in lost revenue, which could be worse.

The data patterns that emerge from the box plot (Figure 9) provide critical insight that will be of interest to innovators and technology implementers in making risk assessments for technology. Consider the following grouping of scores (Table 8).

32

Table 8. Prediction Tool Scores and Risk Total Percentage Score

Unsuccessful

Middle of the Road

Successful

Percentage Successful

0–30%

4

0

0

0%

30–70%

4

5

11

73%

70–100%

0

0

4

27%

Total

8

5

15

It is the middle band that is of most interest, since it contains 11 of the 15 successful outcomes. Also, this is the band where the outcome is unknown and risk assessment is required. Businesses that are risk averse will miss out on 73 percent of innovations, and possible lose their competitiveness. At the same time, businesses with higher risk tolerance and drive for innovation will be more inclined to pursue technologies with lower scores, even when there is possibility of failure. Organization culture, drive to innovate, and ability to absorb occasional financial loss are all factors that need to be evaluated during a risk assessment.

In summary, one of the major results of the predictive model is that as a particular technology matures, it will yield higher performance scores.

Technologies with high performance scores may no longer be cutting edge, but provide a low-risk and potentially high-reward alternative to an unproven technology. However, not all organizations will limit their technology search to mature, proven technologies. The predictive model is especially valuable in identifying key characteristics of emerging technologies that provide attractive risk-reward tradeoffs. Therefore, using the predictive model to choose among alternative technology investments is the first step in leveraging technology to improve construction productivity.

33

A Final Validation of the Tool: The Field Trials A final leg of the validation was to use the tool to assess the RFID/GPS automated material tracking system and compare the tool’s prediction with the team’s actual experience of using the system during the field trials. Overall, the system scored 42 percent. Does this mean that the technology should not be implemented on a construction jobsite? No, it does not. The RFID/GPS automated material tracking system currently falls within the middle band, which represents the area of emerging technologies. Indeed, the system is currently an emerging technology. No comprehensive commercial package for such a system currently exists; researchers developed the algorithms themselves to incorporate the RFID and GPS systems Specific characteristics of the technology that prevent it from scoring higher include its relative lack of maturity, since it has only been tested in a limited field trial effort. Currently, open standards or a dominant vendor do not exist for the system, which represent a risk in the technology’s legacy. These issues are not a criticism of the technology itself but are rather characteristics of most new technologies before they become commercial products. A construction firm that invests in the implementation of such new systems does face risk, but also faces the potential of high rewards and strong competitive gains.

An investment strategy as part of a well-balanced investment portfolio in new technologies from an organization of company perspective is warranted.

34

5 Prioritized List of Technologies Any prioritized technology list is a “snapshot in time.” RT 240 presents two lists of technologies based on the discussion and analysis presented in the preceding chapter. The first is a list of successful technologies that perhaps have yet to be fully exploited by our industry (Table 9).

Relatively new but already widely accepted technologies such as Global Positioning Systems (GPS), orbital welders, robotic total stations, or cellular phones are not included. The second list is one of emerging promising technologies (Table 10). According to the research analysis, these have a roughly 50/50 chance of being successful in any particular organization on any particular project, but show significant potential for impact and have some proven feasibility. The associated predictive model scores represent the range that the respective technologies scored on the predictive model. These scores should be interpreted with a healthy degree of scepticism, and a company or organization should incorporate its own risk analysis before implementing any of the shown technologies.

Table 9. Successful But Not Yet Fully Exploited Technologies in 2008 Predictive Model Score

Technology Building Information Models (BIM) compiles information and allows the user to create a virtual model of a given structure Stakeless Earth Moving GPS-based machine guidance system or grade control system 4D Visualization 3D simulation of a construction schedule Glass Fibre Composites Glass fibre material that is stronger, safer, longer-lasting, and impervious to hazards 3D Laser Scanning Measuring devices that can create 3D images of a jobsite

35

66–81%

Table 10. Emerging Promising Technologies in 2008 Predictive Model Score

Technology Automated Rebar Tier Holds the crossed reinforcing bars, feeds the tie wire, winds, cuts, and ties in one action.

Wearable or Ultra-mobile Computers Mobile personal computer that allows field personnel to access schedule data, email, drawings & specifications anywhere on the job site Next Generation Wireless (WIMAX) WiMAX ,the Worldwide Interoperability for Microwave Access, is a telecommunications technology aimed at providing wireless data over long distances on the scale of multiple miles RFID/GPS Based Materials Tracking Intergrated RFID and GPS material tracking system that provides detailed location information for materials anywhere outdoors on a jobsite.

42–60%

Security Mobile Locks Small devices equipped with a GPS locator that can be attached to containers, equipment and materials enabling the user at any time to determine its location.

Automated Learning Computer-based training systems Infrastructure Sensors Wireless sensors capable of monitor fire detection, elevator safety, security, deflection, corrosion of steel reinforcement and building elelements.

Augmented Reality Computer Aided Drawing Generation of virtual design spaces that may be used not only by the design function, but also to support the development and execution of construction plans.

Readers might ask why RFID/GPS based materials tracking does not score higher. As discussed in the previous chapter, it does demonstrate great promise and potential impact. It is technically feasible and can improve productivity. Currently, however, there is no industry standard, no dominant vendor, and limited proven field applications.

36

6 Conclusions and Recommendations This research did characterize and quantify the impact that changes in technology have had or can have on construction productivity.

However, the extent that a technology can improve productivity is often dependent on factors that do not necessarily relate to the technology itself. An organization’s or a project’s current work processes and what it means to successfully implement a new technology is critical. Most new technologies require work processes to be changed in order for their greatest impact on improving productivity to be realized. Furthermore, a willingness to change work processes requires total support from management. Other issues influence a technology’s potential success including its synergy with existing technologies, the end users’ perception about its ease of use, and the ability of the new technology to integrate with existing ones.

Other conclusions can be made about leveraging technology to improve construction productivity: • Changes in equipment, material, and the automation and integration of project work functions are related to improvements in construction productivity. Changes in the functional range, energy output, and automation of construction equipment are historically related to improvements in labor productivity. In regards to construction materials, changes in the unit weight, modularity, and the improved ability to install materials under harsh conditions were found to be significantly related to productivity improvement. Finally, the integration and automation of project information systems were significantly related to productivity improvement too.

• The use of automated materials identification and localization technologies improves materials management, which is a vital function for maintaining productivity at the construction workface. The study’s field trials showed

37

that not only did improvement in materials tracking improve productivity at the construction workface and the retrieval of materials in a lay down yard area, but also improved the predictability and reliability that materials would be available when needed, which had many intangible effects on the projects. Reengineering existing construction processes is often needed to maximize a technology’s potential benefit.

• Understanding how past technologies improved construction productivity can be used as a predictor of how future technologies might do the same. Utilizing results of the study’s historical data analyses as well as the analyses from other studies, RT 240 developed a tool to predict a technology’s potential success at improving productivity.

Finally, organizations should approach their adoption of new technologies in the same manner they would approach developing any well-balanced investment portfolio. Should an organization begin investing in every new technology that has potential to improve its productivity? There is risk with investing in some new technologies whose benefits have not been measured and proven in actual field processes.

Likewise, if an organization waits to invest in a technology after it has been widely proven to be successful, the risk is significantly reduced, but the technology is likely conventional and already widely adopted by the competition. As a result, a significant competitive advantage of adopting a new technology is lost. Furthermore, if all organizations only invested in conventional, well-proven technologies, it is unlikely that significant step changes in construction industry productivity will occur.

38

Appendix: Map Usage During Field Trials

39

Notes

41

Leveraging Technology to Improve Construction Productivity Research Team Sergio Arantes, Petrobras Ron Bond, Tennessee Valley Authority, Co-Chair * Carlos Caldas, The University of Texas at Austin † Robert Chapman, National Institute of Standards and Technology Steve Davis, WorleyParsons Limited † Shrikant Dixit, Bechtel, Co-Chair * Paul Goodrum, University of Kentucky David Grau Torrent, The University of Texas at Austin * Carl Haas, University of Waterloo † David Heaton, CCC Group, Inc. Daniel Homm, University of Kentucky Leandro Iglezias, Petrobras Richard Jackson, FIATECH Sylvia Kendra, Smithsonian Institution Victor Puccio, URS Saiedeh Razavi, University of Waterloo Thomas Royster, J. Ray McDermott Kamel Saidi, National Institute of Standards and Technology Brian Schmuecker, U.S. Department of State Wayne Sykes, Aker Solutions Todd Vandenhaak, Pegasus Global Holdings Jordan Yeiser, University of Kentucky Dong Zhai, University of Kentucky Past Members Mike Alianza, Intel, former Co-Chair Sean Rooney, Fluor, former Co-Chair † George Stevenson, Bechtel * Principal authors † Joint CII–FIATECH participants Editor: Rusty Haggard

Construction Industry Institute® The University of Texas at Austin 3925 W. Braker Lane (R4500) Austin, Texas 78759-5316 (512) 232-3000 FAX (512) 499-8101