Considering the M3 approach: measure, model, and manage
By Patrick C. Suermann, Ph.D., P.E., LEED AP, Texas A&M University; Dean Wilt, Booz Allen Hamilton; and Pete Kelsey, Carahsoft
The U.S. Department of Defense (DoD), the largest real estate property owner in the world, is struggling to maintain its enormous infrastructure portfolio in support of both current and anticipated mission requirements. Though the department spends about $20 billion annually on facilities sustainment, restoration, and modernization, a recent DoD report rated 32 percent of its facilities worldwide in “poor” or “failing” condition. Compounding these issues, the methods DoD uses to collect and analyze infrastructure data to make portfolio management decisions are labor-intensive and costly.
DoD manages more than 24.9 million acres of land worldwide, occupying 276,770 buildings comprising more than 2.2 billion square feet. These locations also contain 178,113 structures (e.g., towers, storage tanks, piers, and wharfs), and another 107,092 linear structures (e.g., runways, roads, pavement, fences, and electrical distribution lines). To inspect and determine what infrastructure is in need of repair, upgrade, or replacement, DoD currently sends out inspection teams to physically observe and evaluate each building, tower, road, pipeline, etc. The team manually gathers assessment data and often stores it in disconnected systems, thus preventing the sharing of information for advanced analytics or enterprise-wide analysis. Errors creep into the data through manual input or the subjective assessments of individual inspectors. Time-consuming methods of physical inspection also mean that years—and sometimes decades—may pass between inspections.
Most problematic is that current practices for inspection and data collection do not generate the insight necessary to guide effective decision-making for infrastructure investments. With much of the data stored in silos, it is difficult to meet the demands for timely, accurate, and integrated perspectives that drive well-informed investment decisions. As a result, decision-makers often focus narrowly and inefficiently on their most immediate needs, because they do not have the tools or information that can provide a strategic view of how best to optimize infrastructure in support of current and future missions.
DoD’s challenges should resonate with anyone who operates and maintains facilities and infrastructure. This kind of “business as usual” approach is not sustainable. The challenge is this: How can leaders make informed and effective infrastructure investment decisions—decisions consistent with an enterprise-wide infrastructure management strategy—in today’s budget-constrained environment?
Any physical structure—be it a building, runway, or large facility with design and construction predating the digital age—can be digitized. The vast majority of DoD facilities and infrastructure predate CAD/BIM as such no digital data exists. Owner/operators can gain the insight they need for objective, data-driven investment decisions by incorporating commercially available technologies and solutions into their infrastructure inspection, analysis, and decision-making processes. Many organizations are already familiar with technologies such as unmanned vehicles and sensors, which they use to support other mission areas. The key is understanding how to apply the technologies to address modern infrastructure challenges and needs.
Many commercial companies have already begun using some of these tools to inform their infrastructure investment decisions. The digitization process and the technology involved is quickly revolutionizing the operation, maintenance, security, and safety workflows and protocols of these assets.
- This article is part of USGIF’s 2019 State & Future of GEOINT Report. Download the PDF to view the report in its entirety.
The M3 Solution
Consider the M3 approach: measure, model, and manage. The measure phase calls for the digitization, or capture, of the asset. Numerous sensor types are available for this phase including LiDAR, SONAR, Ground Penetrating RADAR (GPR), Mechanical Resonance Imaging (MRI), and photogrammetry. All of these sensors can be ground-based, drone-based, manned, or unmanned.
The data provides a lot of value, but the creation of a solid model, the model phase, can provide even more value in terms of visualization, simulation, and analysis.
Once the measurement data are captured and a model exists, stakeholders have a digital twin of the asset that can be repurposed for any number of scenarios including safety, security, outreach, and education. This is the management phase: managing assets virtually. Later, virtual 4D and 5D analysis can be run to determine the time and cost required to make improvements and repairs. “What if?” scenarios of many types then become possible to evaluate.
Use Case No. 1: The USAFA Cadet Chapel
Designed in the late 1950s and dedicated in September 1963, the U.S. Air Force Academy (USAFA) Cadet Chapel in Colorado Springs, Colo., is a stunning structure featuring 17 identical spires that soar 150 feet into the air and a 99-foot clearance inside the iconic chapel. Able to hold five simultaneous services for a variety of faiths, it is a modernist architectural gem and Colorado’s most visited manmade structure.
Unfortunately, 55 years of exposure to the elements will take its toll on any building. The chapel’s concrete foundation, for example, has been damaged by annual freeze-thaw cycles, and the building is experiencing water infiltration. Additionally, the original gaskets and seals system were value engineered in the original construction and the building has subsequently leaked since its commissioning.
To determine what other repairs might be needed and how best to communicate the need for renovation to key stakeholders, USAFA decided to assess the existing state of the chapel by documenting the structure in an entirely new way.
To tackle this task, the academy partnered with Autodesk, whose ReCap reality capture software is able to digest data from multiple sources and generate a single, photorealistic 3D model capturing every detail of a structure. Working with the 21st Civil Engineer Squadron from nearby Peterson Air Force Base, the measurement phase was completed using ground- and UAV-based LiDAR and photogrammetric technology.
The modeling phase involved the creation of a Building Information Model (BIM) using Autodesk Revit software.
Now that the chapel is modeled in Revit, the possibilities in the management phase are nearly limitless. Whole-building envelope energy and lighting analysis may be performed. Exterior wind studies can be conducted. Computational Fluid Dynamics analysis can perform more advanced studies to inform the best HVAC design to save energy and ensure the comfort of chapel occupants. A wide range of structural analyses are possible, including a new wind load simulation feature that would be useful in studying the unique geometry of the chapel. Direct links from the model to visualization programs enable unmatched visualization and animation possibilities. The ultimate result is a digital twin of the chapel that is being used to effect repairs, and that will serve indefinitely as a baseline dataset to measure change, as well as a successful proof of concept to show the greater Air Force new ways to manage assets.
Use Case No. 2: The Rapid Airfield Damage Assessment System
DoD is well aware of the vulnerability of its airfields. Operating and maintaining airfields in combat situations is mission-critical in terms of maintaining air supremacy. As such, the ongoing Rapid Airfield Damage Assessment System program is all about assessing damage, identifying and mitigating any unexploded ordnance (UXOs), and accomplishing repairs as quickly as possible. Performing these tasks remotely and robotically is paramount for the safety of the technicians. Finally, time is the critical element. All of this must happen within hours so the airfield can resume operations.
As different as this scenario may appear to the USAFA Cadet Chapel, the M3 workflow still applies. Establishing a baseline dataset of the intact airfield, assessing damage, and identifying UXOs is the measurement phase. Determining the most efficient path to a repaired and operational airfield can be done from both measurement and model data. Updating and maintaining the existing airfield model, which includes the repairs, is the management phase. The Air Force has heavily invested in technology that helps eliminate human physical inspection and replaces it with reality capture and artificial intelligence-based change detection. The future of airfield repair will be safer and more rapid due to this technology.
Use Case No. 3: Glen Canyon Dam
Glen Canyon Dam (GCD) in Page, Ariz., is operated and maintained by the U.S. Bureau of Reclamation (USBR). Designated national critical infrastructure by the Department of Homeland Security, it is a secure facility that operates 24/7/365. As GCD is a hydroelectric dam it not only manages the water supply to much of the Southwest, it also provides electricity to the region.
Designed and built in the ’50s and ’60s, little to no digital data existed of the facility. Operations, maintenance (O&M), and security were conducted as efficiently as possible using the analog data of the period. In 2016, USBR decided to embark on a proof of concept project to create a digital twin of the facility. Confidence was high that the resulting model would provide value well beyond the original project scope focused on O&M. The measurement phase used LiDAR, SONAR, and photogrammetry to create the digital twin. A solid CAD/BIM-based model was then created.
The data provided value almost immediately. Moving large, heavy pieces of equipment in and out of the facility safely and efficiently could be modeled virtually. Would the equipment fit? How long would it take? How much would the upgrade/repair cost? What was the state of silt buildup on the upstream side of the dam related to the intakes? Insight into these important questions could all be better determined and planned for virtually. The final phase of the project will be to create a virtual facility that can be used to run security-based scenarios such as evacuations, terrorist events, and even outreach and educational efforts as the facility has high tourist traffic.
Diving Deeper into the M3 Process
The main elements underlying the M3 process for improving infrastructure assessments and decision-making include:
- Innovative Autonomous Data Collection Platforms: DoD and commercial organizations can accelerate data collection by using UAVs, ground vehicles, submersibles, and other autonomous platforms equipped with sensors. In addition to gaining easier access to remote or hazardous locations, these platforms can also fill gaps in current data collection. DoD could also repurpose existing platforms such as satellites to enhance data collection.
- Multi-spectral Sensors: Deploying video and photographic imaging as well as infrared, topographical, and LIDAR sensors on fixed assets or autonomous platforms to collect real-time imaging and infrastructure data has proven effective. Advances in storage and computing power have enhanced the ability of small, sensor-carrying platforms to collect and deliver valuable data.
- Powerful Cloud Computing: Organizations already collect large stores of data that can inform infrastructure management. They can also take advantage of public or open-source datasets as well as enhance existing data with new sensor data from unmanned systems and other sources. The maturation of cloud computing enables organizations to store, access, share, and manage massive amounts of data mined from these varied sources.
- Advanced Analytical Solutions: Modern analytic tools can apply sophisticated algorithms and models to parse through large amounts of data to detect anomalies and trends and conduct multi-dimensional analysis of potential problems, solutions, and costs.
- Visualization: Modern decision-making tools can integrate and present complex analytic findings in dashboards and other displays, providing decision-makers with a realistic picture of the infrastructure situation in a 3D/4D virtual or alternative visual reality (e.g., simulations, holograms, and multidimensional analysis). This enables them to clearly analyze and compare investment trade-offs across their entire infrastructure portfolio.
Applied together, these advanced technologies offer many opportunities to accelerate the data collection process, generate added value or return on investment for existing data, and create new efficiencies and solutions that go far beyond the 1950’s-style inspection methods currently in place. For example, UAVs and other data collection platforms, combined with next-generation analytics and data visualization, can anchor a more efficient, accurate, and timely assessment approach—and at a substantially lower cost than physical inspection teams. Infrared UAV sensors enable surveys of an entire infrastructure portfolio, efficiently pinpointing areas of concern.
Equipped with LiDAR to establish a building’s 3D structure and GPS to supply coordinates of a particular building, UAVs can leverage infrared capabilities for surveillance of larger groups of buildings in a single flight. Thus, as opposed to inspecting buildings individually, airborne infrared sensors allow for an expedited analysis of potential structural issues in a multiple-structure or base-wide facility portfolio.
The advantages of using unmanned vehicles to inspect buildings also hold for inspections of other types of structures, such as nuclear facilities, airports, ports, pipelines, towers, and bridges. This allows teams to conduct surveys more rapidly and on a more regular basis. The inspection processes are repeatable, and the sensor data are reliable, measurable, and shareable, while the automated inspection reporting offers decision-makers on-demand access to an objective, enterprise-wide view of their infrastructure portfolios, significantly shortening the inspection cycle. It should be noted that physical teams are still necessary—for example, to inspect infrastructure where sensor data have indicated a problem, such as a steam pipe leak, excessive corrosion, or unwanted moisture. But the overall reduction in the need for physical inspections to discover these issues will reduce costs while increasing worker safety, particularly when inspections cover large expanses or include rugged, difficult-to-reach locations and confined spaces.
Implementing M3 for Better Decisions
Capitalizing on M3 will require a new approach to inspecting infrastructure, analyzing the data, and presenting the results to decision-makers. The goal isn’t simply to collect more data, but more importantly, to replace outdated practices with more powerful, cost-efficient methods, while also leveraging existing data sources. To extract maximum value from the collected data, the new approach will integrate the diverse datasets to facilitate rigorous analytics and present a comprehensive, enterprise-wide view of the infrastructure portfolio.
The first step involves developing a diagnostic of existing data and client needs: What data does the organization already collect? What additional data do decision-makers need to effectively assess their infrastructure and plan investments? Data collection and analysis represent a complex and challenging undertaking due to a number of factors such as lack of relevant data, an abundance of data, the high cost to collect data, data anomalies, multiple data systems, and gaps that limit the ability to conduct integrated analysis.
With this insight, the next set of activities focuses on filling gaps and accelerating data collection using innovative platforms and sensors, open-source and untapped data sources (e.g., weather), and techniques such as geo-tagging to enhance data.
We then establish the processes and controls for connecting accurate, timely data. Accurate data provide the foundation for high-end analytics and the development of alternatives. It is essential for identifying efficiencies, tracking trends, and making process, facility, and equipment adjustments, as well as for demonstrating and quantifying actual cost savings, efficiencies, and schedule improvements.
Identifying the right datasets and knowing what to analyze requires staff with expertise and knowledge in facilities, public-private partnerships, resource management, and the organization’s mission. This expertise, when combined with strong communications skills, an ability to work across functional boundaries, and an understanding of the links and synergies of data help to achieving operational, cost, and schedule goals.
The final set of activities focuses on providing results by strengthening decision-making through modeling, simulation, and machine learning. Organizations can see a significant return from their expanded data collection by applying advanced decision-making tools and techniques, which allow for complex sets of data to be analyzed faster and more cost effectively. These tools can analyze enormous datasets to highlight the anomalies and trends that vary from the rest of the set. In addition, the tool’s capacity for multi-dimensional analysis enables it to study more factors at one time, allowing the analyst to understand how each variable relates to the others.
M3 workflows help organizations pinpoint and resolve problems within a large infrastructure portfolio with limited time and resources. Moreover, this activity creates an integrated analysis of alternatives that generates clear recommendations and enhanced decision-making consistent with the infrastructure management strategy.
 Statement of Mr. Pete Potochney, Performing the Duties of Assistant Secretary Of Defense (Energy, Installations and Environment) Before the House Appropriations Committee Subcommittee on Military Construction, March 3, 2016, p. 5.
 “Base Structure Report – Fiscal Year 2015 Baseline: A Summary of the Real Property Inventory,” pp. 7-8.
Headline Image: U.S. Air Force photo/Staff Sgt. Andrew Park