Department of Defense uses data for predictive logistics planning
An Air Force sergeant logs data into the Cargo Movement Operations System.
Ministry of Defense photo
oklahoma city — “Logistics wins wars” is an old adage, but its applications have evolved from logbooks and spreadsheets to data streams and analytics. The Department of Defense now wants to leverage these new practices to predict the future.
Christopher Roman, Assistant Secretary of Defense for Sustainability, is working with a variety of platforms to better understand what tools need to be used to turn data into actionable logistics decisions. We described predictive logistics as a way to utilize the generated data. This is no shortage for the Department of Defense.
“It’s all about restoring readiness and getting the ability to meet demand within the theater of operations closer to the point of need,” Roman said at the Defense Industries Association’s recent National Logistics Forum.
Gen. Charles Hamilton, commander of the Army Materiel Command, said precision and predictive persistence “is not just about knowing when, where, and how much ammunition and other things are being used by the unit.” [or] It doesn’t just mean water or specific parts, it also means knowing exactly where, when and how much of the product will be needed in the future. And the important thing is to come forward.”
The ability to predict is also the ability to stay one step ahead of the enemy, which requires having a “data-driven decision-making process” to “go into the next battle,” Hamilton said.
A data analytics “crystal ball” takes all the data collected across government and industry and analyzes and interprets meaningful patterns. Hot-button solutions like artificial intelligence and machine learning are helpful tools, but they need to be recognized as such, participants in the forum panel suggested. Basic inputs are dependent data.
Kevin Gaudette, a retired Air Force colonel and senior vice president of integrated analytics and support at LinQuest, says AI and machine learning, like optimization and simulation, are “tools and data-dependent.” Masu. When we throw these buzzwords around, everyone gets excited and starts asking for things they don’t even know what they want. ”
The question should be “What are you trying to do?” He said. “Let’s start by understanding what it is. Because [AI and ML] It may not be the answer. In many cases, this is not the case. ”
But sometimes it does.
Marine Corps Maj. Gen. Keith Reventlow, commander of Marine Corps Logistics Command, called artificial intelligence and ChatGPT “amazing tools.” Although it may be associated with routine reporting fraud, the concept of “looking at all sources of information on the Internet” [a] The theme of “compare everything and give an answer” could be revolutionary for logistics, he said.
“When you think about sustainability, when you understand weapon systems, when you understand the predictability of maintenance in context, you can do similar things, like predict time to failure, and give commanders the ability to change that. “What if we were given the option?” he said. before it actually breaks. I think we’re working on all sorts of opportunities to understand how to leverage data using these tools. ”
AI and machine learning are just some of the tools in a wide range of projects and research underway across the Department of Defense to better understand how data can be used.
One such initiative is the Defense Logistics Agency’s Joint Additive Manufacturing Model Exchange (JAMMEX). It is a tool that integrates technical data packages and allows users to download and print models from multiple sources through a single system.
Adaryl Roberts, chief information officer for information operations at the Defense Logistics Agency, highlighted a program called the Digital Sustainment Platform that is built on JAMMEX. “Not only can we take the technical data package, but we can also create a single platform that is available to all engineers. We can bring digital twins across the department to increase efficiency across warfighter commands and services. You can place it.
Roberts said DLA’s other initiatives include digital business transformation and warehouse modernization projects. These efforts have bought up what he called 20 years of technical debt, leaving the department “left behind in terms of flexibility and the ability to be nimble with technology,” he said.
The Army launched a cross-logistics team last year, and Hamilton said, “It’s integrated with everything else we do, so it’s never going to go away.” Part of the team’s “very narrow problem set” is accuracy and predictive persistence, he said.
Air Force Materiel Command launched a digital materiel management initiative last year. This is an initiative that leverages digital tools, structured data, and security to integrate and adopt digital methodologies across the entire functional lifecycle, from invention to disposal.
“We created an organizational structure, aggregated data, identified all risks, and built a team to mitigate them,” said Jim Sutton, senior director of strategy at Shipcom Wireless, at a recent kickoff event for the initiative. He said there were conversations about the idea of establishing a -The conversation around Zero Trust is completely non-existent.
Zero Trust is the Department of Defense’s security framework premised on the idea of ”never trust, always verify.” There is well-documented hesitancy between government and industry when it comes to trust when it comes to data sharing, creating a further barrier to effectively leveraging data analytics.
While there is a lot of “fundamental uncertainty” around data sharing as industry and government intersect on the delivery of capabilities, Sutton said the main entry point from a service leadership perspective is the White House’s 2022 The aim should be to encourage implementation of the zero trust memorandum. Government agencies are required to meet certain cybersecurity standards by the end of fiscal year 2024.
“If that were in place, virtually 90 per cent of any discussion of risk would disappear,” he said. “The result is a focus on where risk really matters in terms of having actionable, decision-worthy information at the point of use.”
Aaron Jaffe, Director of Supply Chain and Logistics at Palantir Technologies, said that data collaboration between industry and the Department of Defense is “an increasingly important input to the broader industrial base as a limiting factor in how the Department of Defense operates.” We need to think about it as a community.” It will enable us to integrate and collaborate with allies and partners in every decision we make. ”
The question is how to better understand the gaps and seams and the technology that protects security while enabling seamless interoperability “across a much larger community than we’ve seen before.” He said it would be different.
The gap poses “immense challenges” but is not unprecedented, he said. Operation Warp Speed, the COVID-19 vaccine acceleration effort, and aid to Ukraine are examples of industry and government working together on a fast-track timeline and “moving at the pace of conflict.”
He suggested that at the core of making these efforts possible, and what the industry needs now, is a clear sense of mission. It builds on its open and interoperable approach. By leveraging existing programs and new technologies, we can increase the pace of change across the Department of Defense. ”
One industry representative said that while reliable data is critical, the Pentagon also needs to get used to dealing with dirty data.
Justin Wolf, Systecon’s chief technology officer, said perfect data will probably never be achieved and while it is a noble and necessary pursuit, the sector must accept some risk. Ta.
“What we really need to do is actually start doing the analysis,” he said. He added that the idea of dirty data is “appealing” but not an obstacle, saying that while no major acquisition program has perfect data, “data as dirty as it is today” can be used to The data suggested that a complete mission possible rate could be predicted. His 3% of the goal achieved.
Pure data is ideal, but “it’s not an excuse not to do analysis,” Wolf says. “So the idea that we can’t start anything until we have perfect data is nonsense. When we think about competitive logistics, we’re planning for the unknown. …So why maintenance data and supply transactions? Do we really need to know the full history?”
There’s definitely some ambiguity, he said, “but we can throw aside the idea of ’I have to know everything before we start’ and just use analytics and learn more about analytics, our systems. For…and [accept] that [there are] Sometimes you don’t know completely, but you go from there, apply the risk, understand the risk, and make decisions. ”
Gaudette said this for the first time. “For a long time, we’ve had an integrated strategy document and an integrated architecture, but the data hasn’t been collected yet. We’re going to end up going in a million different directions. So our company There is a lot to like about it, but to date there is still no integrated solution.”
There may not be a definitive solution, but the Pentagon is looking, and the most important thing is to “start,” Sutton said.
Leveraging data to make informed, predictive decisions means putting the tools in the hands of operators, and “the more you do it, the better you get at it,” he said. . …And the more we encourage employees to try using it in spaces that help them do their jobs, the more prevalent it will be and the easier it will be adopted. ”ND
topic: Logistics/Logistics/Maintenance
