Back to news

July 8, 2019 | International, Other Defence

How the Pentagon can improve AI adoption

By: Graham Gilmer

The excitement of artificial intelligence today is like the space race of the 1960s, when nations were in fierce competition. Now, the United States is in first place. But continued leadership is not a given, especially as competitors, namely China and Russia, are making significant investments in AI for defense. To maintain our technological advantage, safeguard national security, and lead on the world stage, we have an imperative to invest strategically in AI.

The successful and widespread adoption of AI requires the United States take a human-centric and technologically innovative approach to using AI to help maintain the peace and prosperity of our nation. As the Department of Defense and Joint Artificial Intelligence Center (JAIC) continue their efforts to accelerate AI adoption, they must address three key components of successful adoption: building trust in AI technology, operationalizing AI technologies to reach enterprise scale, and establishing ethical governance standards and procedures to reduce exposure to undue risk.

Build trust in AI technology

Fear and distrust hold technology adoption back. This was true during the first three industrial revolutions as mechanization, factories, and computers transformed the world, and it is the case in today's fourth industrial revolution of AI. The confusion surrounding AI has led to teams abandoning applications due to a lack of trust. To build that trust, we must prioritize training, explainability, and transparency.

Trust in technology is built when leaders have accurate expectations around what it is going to deliver, mission owners can identify use cases connected to the core mission, and managers understand the true impact on mission performance. Building trust requires that all users, from executives and managers to analysts and operators, receive training on AI-enabled technologies. Training involves not only providing access to learning resources, but also creating opportunities for them to put their new skills to use. In its formal AI strategy, Pentagon leaders outlined extensive plans for implementing AI training programs across the department to build a digitally savvy workforce that will be key to maintaining the United States' leading position in the AI race.

“Explainable AI” also curbs distrust by showing users how machines reach decisions. Consider computer vision. Users may wonder: How can such a tool sift through millions of images to identify a mobile missile launcher? A computer vision tool equipped with explainable AI could highlight aspects of the image that it uses in identification—in this case, elements that look like wheels, tracks, or launch tubes. Explainable AI gives users a “look under the hood,” tailored to their level of technical literacy.

AI technologies must be more than understandable; they must also be transparent. This starts at the granular system level, including providing training data provenance and an audit trail showing what data, weights, and other inputs helped a machine reach its decision. Building AI systems that are explainable, transparent, and auditable will also link to governance standards and reduce risk.

Operationalize AI at the enterprise scale

AI will only be a successful tool if agencies can use AI at the enterprise level. At its core, this means moving AI beyond the pilot phase to real-world production across the enterprise or deployed out in the field on edge devices.

Successfully operationalizing AI starts early. AI is an exciting new technology, but agencies too enamored with the hype run the risk of missing out on the real benefits. Too many organizations have developed AI pilot capabilities that work in the lab but cannot support the added noise of real-world environments. Such short-term thinking results in wasted resources. Agencies must think strategically about how the AI opportunities they choose to pursue align with their real-world mission and operations.

Leaders must think through the processes and infrastructure needed to seamlessly extend AI to the enterprise at-scale. This involves building scalable infrastructure, data stores and standards, a library of reusable tools and frameworks, and security safeguards to protect against adversarial AI. It is equally important to prioritize investment in the infrastructure to organize, store, and access data, the computational needs for AI (cloud, GPU chips, etc.), as well as open, extensible software tools for ease of upgrade and maintenance.

Establish governance to reduce risk

Governance standards, controls, and ethical guidelines are critical to ensuring how AI systems are built, managed, and used in a manner that reduces exposure to undue risk. While our allies have engaged in conversations about how to ensure ethical AI, China and Russia have thus far shown little concern for the ethical risks associated with AI. Given this tension, it is imperative that the United States maintain its technological advantage and ethical leadership by establishing governance standards and proactive risk mitigation tactics. To this end, in May, three Senators introduced the bipartisan Artificial Intelligence Initiative Act, which includes provisions for establishing a National AI Coordination Office and national standards for testing AI algorithm effectiveness.

Building auditability and validation functions into AI not only ensures trust and adoption, but also reduces risk. By establishing proactive risk management procedures and processes for continuous testing and validation for compliance purposes, organizations can ensure that their AI systems are performing at optimal levels. Governance controls and system auditability also ensure that AI systems and tools are robust against hacking and adversarial AI threats.

AI could be the most transformative technological development of our lifetime—and it's a necessity for maintaining America's competitive edge. To ensure that we develop AI that users trust and can scale to the enterprise with reduced risk, organizations must take a calm, methodical approach to its development and adoption. Focus on these three areas is crucial to protecting our national security, maintaining our competitive advantage and leading on the world stage.

Graham Gilmer is a principal at Booz Allen who helps manage artificial intelligence initiatives across the Department of Defense.

https://www.c4isrnet.com/opinion/2019/07/08/how-the-pentagon-can-improve-ai-adoption/

On the same subject

  • As tech startups catch DoD’s eye, big investors are watching

    January 31, 2020 | International, Aerospace, Naval, Land, C4ISR, Security

    As tech startups catch DoD’s eye, big investors are watching

    By: Jill Aitoro SIMI VALLEY, Calif. — Private investors are not yet lining up to back defense startups, but they are paying close attention. Two factors have created an opening that could lure venture capitalists to defense investments: first, a few select venture-backed technology startups are gaining traction; and second, there's been a strategic shift in approach to weapons development from the U.S. Department of Defense, focusing more on information warfare and, as such, software. In the words of Mike Madsen, director of strategic engagement at the Pentagon's commercial tech hub, Defense Innovation Unit: "We're at a significant inflection point right now that will be visible through the lens of history.” Nonetheless, for the tech startups, it's been slow going, as discussed during a Defense News roundtable in California. For the second year, leadership from DoD and the tech community came together to discuss the state of the Pentagon's efforts to attract commercial startups — this time digging into the challenges and opportunities that come with investment in defense development. “We went into this eyes wide open, knowing full well that to the venture community, the math doesn't make sense. Making the choice to contribute to the advancement of artificial intelligence for DoD represented for us more of a mission-driven objective,” said Ryan Tseng, founder of artificial intelligence startup Shield AI. But early on, “we were fortunate to get the backing of Andreessen Horowitz, a top-tier venture fund. They're certainly leaning in, in terms of their thinking about defense technology — believing that despite the history, there might be a way to find an opening to create companies that can become economically sustainable and make substantial mission impact.” Shield AI has raised $50 million in venture funding since 2015, with more rounds expected. Indeed, a few key Silicon Valley investors have emerged as the exceptions to the rule, putting dollars toward defense startups. In addition to Andreessen Horowitz, which counts both Shield AI and defense tech darling Anduril in its portfolio, there's General Catalyst, which also invested in Anduril, as well as AI startup Vannevar Labs. And then of course there's Founders Fund. Led by famed Silicon investors Peter Thiel, Ken Howery and Brian Singerman, among others, the venture firm was an early investor in Anduril, as well as mobile mesh networking platform goTenna. Founders Fund placed big bets on Palantir Technologies and SpaceX in the early days, which paid off in a big way. Some of the early successes of these startups have “done an excellent job of making investors greedy,” said Katherine Boyle, an investor with General Catalyst. “There's a growing group who are interested in this sector right now, and they've looked at the success of these companies and [are] saying: ‘OK, let's learn about it.' ” Take Anduril: The defense tech startup — co-founded by Oculus founder Palmer Luckey and Founders Fund partner Trae Stephens — has raised more than $200 million and hit so-called unicorn status in 2019, reaching a valuation of more than $1 billion. As the successes piled up, so did the venture capital funding. According to Fortune magazine, those investors included Founders Fund, 8VC, General Catalyst, XYZ Ventures, Spark Capital, Rise of the Rest, Andreessen Horowitz, and SV Angel. “I started my career at Allen & Company investment banking. Herbert Allen, who's in his 80s, always said: ‘Hey, you should run into an industry where people are running away,' ” said John Tenet, a partner with 8VC as well as a co-founder and vice chairman of defense startup Epirus. “There's so much innovation occurring, where the government can be the best and biggest customer. And there are people who really want to solve hard problems. It's just figuring out where the synergies lie, what the ‘one plus one equals three' scenario will be.” Also attracting the attention of Silicon Valley investors is the growing emphasis by the Pentagon not only on systems over platforms, but software over hardware. Boyle described the shift as the “macro tailwind” that often drives innovation in a sector. Similar revolutions happened in industrials and automotive markets — both of which are also massive, global and slow-moving. That emphasis on tech, combined with some recent hard lessons, also provides a glimmer of hope that the typical hurdles associated with defense investments — lengthy procurement cycles and dominance by traditional manufacturers, for example — could be overcome. Consider U.S. Code 2377, which requires that commercially available items be considered first in procurement efforts, said Anduril's Stephens. He also noted court decisions in lawsuits filed by SpaceX and Palantir, which ultimately validated claims that defense agencies had not properly ensured a level playing field for major competitions. “These types of things are now at least in recent memory for Congress, and so they have some awareness of the issues that are being faced,” Stephens said. “It's much easier now to walk into a congressional office and say, ‘Here's the problem that we're facing' or ‘Here's the policy changes that we would need.' There are also enough bodies like DIU, like In-Q-Tel, like AFWERX, like the Defense Innovation Board, like the [Defense Science Board] — places where you can go to express the need for change. And oftentimes you do see that language coming into the [National Defense Authorization Act]. It's part of a longer-term cultural battle for sure.” For now, all these factors contribute to the majority of skeptical investors' decisions to watch the investments with interest — even if they still take a wait-and-see approach. And that places a lot of pressure on the companies that are, in a sense, the proof of concept for a new portfolio segment. “My fear is that if this generation of companies doesn't figure [it] out, if they don't knock down the doors and if there aren't a few successes, we're going to have 20, 30 years of just no investor looking around the table and saying we need to work for the Department of Defense,” Boyle said. “If there aren't some success stories coming out of this generation of companies, it's going to be very hard to look our partners in the eye and say: ‘We should keep investing in defense because look at how well things have turned out.'” https://www.defensenews.com/smr/cultural-clash/2020/01/30/as-tech-startups-catch-dods-eye-big-investors-are-watching/

  • 4 questions with NATO on its unmanned tech test

    October 29, 2019 | International, Naval

    4 questions with NATO on its unmanned tech test

    By: Martin Banks BRUSSELS — As militaries around the world invest in advanced technology, the need to test the capabilities of new systems for military operations is critical — both to ensure the training of personnel as well as the effective integration with existing platforms. Dozens of unmanned underwater, surface and air vehicles from NATO countries gathered in Portugal in September for Exercise REP (MUS) 19 to do just that: test technological advances in unmanned maritime systems networks. Defense News recently received details about the exercise from a NATO official. In a nutshell, what is Exercise REP (MUS) 19? REP (MUS) 19 was built on the 10th annual Portuguese underwater exercise Recognised Environmental Picture (REP), with support from NATO's Maritime Unmanned Systems (MUS) initiative, the NATO Centre for Maritime Research and Experimentation, and the University of Porto's Laboratory for Underwater Systems and Technology. The NATO Maritime Unmanned Systems Initiative (MUSI) was launched in October 2018 to promote capability development and interoperability in the field of maritime unmanned systems. What did NATO hope to learn from this exercise? The focus of REP (MUS) 19 was on technological and procedural interoperability. Participating nations tested the integration and coordination of activities between multiple unmanned systems from allied nations in the three domains — above the water, on the water and underwater. This was the first time that so many NATO nations had the opportunity to test together the effectiveness of systems, concepts, techniques and procedures related to maritime unmanned systems, ensuring they can work seamlessly together, bringing together dozens of unmanned underwater, surface and air vehicles for maritime operations. Who participated? The systems were from the Portuguese Navy, as well as from the NATO Centre for Maritime Research and Experimentation, from Belgium, Italy, Poland, Turkey, the United Kingdom, and the United States. During REP (MUS) 19, participants from naval forces, from industry and from academia jointly contributed assets to the operational demonstrations and worked together to test new technological advances and procedures for maritime unmanned systems in real-life operational scenarios. What is the potential for unmanned systems among allies? New maritime unmanned systems technologies can be a game-changer in countering multiple threats in the maritime domain. Using maritime unmanned vehicles can help effectively counter new submarines armed with more powerful weapons. They can also prevent military personnel from moving into risky situations in countering threats like sea mines. https://www.defensenews.com/training-sim/2019/10/28/4-questions-with-nato-on-its-unmanned-tech-test/

  • KC-135s and C-130Js are the next aircraft to enter the Air Force’s data-driven maintenance program

    October 29, 2018 | International, Aerospace

    KC-135s and C-130Js are the next aircraft to enter the Air Force’s data-driven maintenance program

    By: Jeff Martin The KC-135 refueling tanker and the C-130J airlifter will be the next two aircraft to be part of the Air Force's predictive maintenance program, part of the service's effort to do maintenance before airplanes break--and improve mission capable and availability rates, according to Air Mobility Command's logistics director. "The bottom line is to get to where we are scheduling all of our maintenance, rather than reacting to the maintenance,' said Brig. Gen. Steven Blaymaier, who oversees Air Mobility Command's logistics, engineering and force protection, in an interview at the 2018 Airlift Tanker Association symposium outside Dallas, Texas. “We want our units to achieve their mission-capable rates on a sustained basis.” According to the latest available data, from FY2017, the KC-135 fleet had, on average, a 74 percent mission capable rate, and the C-130J fleet had a 77 percent mission capable rate. As for the rest of the mobility fleet, by that same data, the C-5M fleet stood at 60 percent, the C-130H fleet was at 73 percent, and the C-17 fleet was at 84 percent mission capable. The concept, known as conditions based maintenance, has already been rolled out to the C-5 fleet within AMC, and the B-1 fleet in Air Force Global Strike Command. It uses algorithms based on reams of data to create models to predict when a part might break, rather than waiting for it to fail. Its a standard practice in the commercial aviation industry, and is now making its way into the Air Force. Blaymaier says the KC-135 fleet will join the program in spring of 2019, and the C-130J fleet will follow in the summer. Blaymaier also added that the other aircraft in AMC's fleet would be joining the program eventually, like the C-17 and KC-10. “They're all in work at their program offices right now,” he said. “What we learn from C-5 will be incorporated into the other aircraft.” Blaymaier also said the Air Force was modelling their effort after Delta Airlines's Tech Ops division's procedures, and that the service was at the beginning or “crawl stage” of the process. He also noted that Delta took “eight years” to get achieve the results they were looking for, and that the Air Force was working on that path. In September, Lt. Gen. Robert McMurray, commander of the Air Force Sustainment Center, told Defense News that the conditions-based maintenance program was critical to increasing the readiness of the Air Force's aircraft. "Given the aging fleet situation that we have, we probably need to be using data better to take care of it — which is a drive toward what most everyone right now is saying is the right way to manage fleet sustainment, which is through condition-based maintenance and data analytics,” he said at the time. Another benefit Blaymaier described was that the service will be able to track maintenance needs by individual aircraft, rather than by a general fleet-wide standard. That could reduce time in depots and increase mission capable rates, a top priority of senior Pentagon leaders. “As we move forward with conditions based maintenance plus (CBM+) and predictive analytics, we'll be able to know by tail number which parts are going to fail on certain aircraft," he said."It'll be much more surgical [and] operational." Blaymaier added that while the transition to conditions based maintenance might be a long journey, it would lead to huge benefits for the Air Force. “Ultimately we want to achieve those aircraft availability standards that we established for each of our fleets that are required to meet our wartime taskings,” he said. https://www.defensenews.com/digital-show-dailies/airlift-tanker-annual/2018/10/27/kc-135s-and-c-130js-are-the-next-aircraft-to-enter-the-air-forces-data-driven-maintenance-program

All news