By David Vergun, Army News ServiceAugust 21, 2018
WASHINGTON -- Artificial intelligence will inevitably be one of those leap-ahead technologies that enable Soldiers to survive on the battlefield and win, said Vice Chief of Staff of the Army Gen. James C. McConville.
McConville spoke here at the National Defender Investigator Association-sponsored Army Science & Technology Symposium and Showcase, Aug. 21.
Artificial intelligence involves software algorithms that enable a computational device to learn as it processes information and to change courses of action automatically based on parameters set by the machine's designer. AI can process data at extremely high speeds as well.
AI programs can also sift through enormous amounts of data, determine which data are important, simplify the data and present options to operators or commanders.
AI could have lots of useful applications, McConville said. For instance, when combined with sensors on vehicles and aircraft, AI could help determine the best time to perform maintenance or replace parts.
If a formation of tactical vehicles are breaching an area suspected of having buried IEDs, it might make sense for the lead vehicle to operate unmanned and autonomously enabled by AI, he said. The same might apply for a lead aircraft formation going into highly contested air space.
In the future, the AI function might control robotic loader and firing mechanisms or provide targeting recognition capabilities to support Soldiers in ground combat.
Still, AI and robots will never replace Soldiers on the battlefield. McConville said with any unmanned or autonomous arrangement, there would always need to be a person in the loop making ethical decisions about whether or not to engage targets. There's also something visceral, he said, about being on the battlefield that decision makers can't experience on a laptop miles away.
Dr. Thomas Russell, deputy assistant secretary of the Army for research and technology, said he was once asked to explain the "third offset" in one word. That word would be "speed," he said.
Whoever acts first on the battlefield wins, he said. "We've got to be more efficient at everything we do and AI will help us do that."
We're in the infancy of AI, he said, predicting AI will be resident in weapons systems of the future. "It's not residing in our platforms currently. What we call AI now is just good data analytics, it's not AI."
The Department of Defense recognizes the importance of AI. Recently, Defense Deputy Secretary Patrick Shanahan signed a memo establishing the Joint Artificial Intelligence Center to accelerate AI research and ultimately AI delivery to the field in all of the services.
Industry is assisting in that effort as well. Retired Navy Capt. Frank J. Michael, an NDIA senior vice president, said NDIA will soon establish an "AI Community of Influence" to create synergies across its 65 divisions, chapters, committees, working groups and affiliates.
Mary Miller, performing the duties of assistant secretary of Defense for Research and Engineering, said AI is one of the DOD's top priorities because it will be a critical component of other DOD priorities such as hypersonics and autonomous ground and air unmanned systems.
Besides developing AI, the other challenge is getting Soldiers to trust AI, she said. They must learn to understand machine thinking and the machine must learn to understand the Soldier's or commander's intent.
MODERNIZATION PRIORITIES AND AI
The Army has narrowed its focus to six modernization priorities, any of which could have an AI element that would better enable them, McConville said. Those priorities include long-range, precision fires; a next generation combat vehicle; future vertical lift; the network; air and missile defense; and Soldier lethality.
Cross functional teams are currently engaged with industry in each of these priority areas, he said. Each of these CFTs will fall under the new Army Futures Command, when it reaches full operating capability next summer.
Through collaboration with industry, the CFTs will shorten the current length of time it takes to take a program of record through the acquisition process and deliver capabilities to the warfighter.
Failure is an option, he added, but failure should occur early on during testing or demonstration rather than once a program is reaching developmental maturity. That way, a change of course is less expensive and the failure itself could be added to the knowledge base.
AFC personnel will embed with industry and academia in the Austin, Texas, metro area while CFTs are co-located and engage with industry throughout the country to find the best and brightest ideas in innovation, he said. "They will provide us with solutions we might not even have considered."
Besides tapping into industry, the Army aims to better tap into its own talent pool of Soldiers, he said. Maybe there's someone with a degree in the AI field, for example, but the Army only knows that person's rank and that he's perhaps an infantryman or logistician. The Army's new Integrated Personnel and Pay System, IPPS-A, which will be fully rolled out within two years, is expected to track that type of talent and skill. Using IPPS-A, commanders who need specific kinds of talent and skill will be able to find the Soldiers who have it.