Bloomberg reports that the Pentagon, rather than going it alone and starting over, is eager to learn from work already done by private companies like Waymo, Uber Technologies and General Motors so it can deploy self-driving vehicles on the battlefield before they appear on busy streets. The Defense Advanced Research Projects Agency, better known as DARPA, has been funding research into autonomous vehicles for years and first sponsored a competition for self-driving cars in 2004.
Military officials see an immediate need for autonomous driving, with more than half of combat-zone casualties attributed to military personnel delivering food, fuel and other supplies.
"You're in a very vulnerable position when you're doing that kind of activity," Michael Griffin, the undersecretary of defense for research and engineering, said during a hearing before lawmakers on Capitol Hill in April. "If that can be done by an automated unmanned vehicle with a relatively simple AI driving algorithm where I don't have to worry about pedestrians and road signs and all of that, why wouldn't I do that?"
The Pentagon has a nearly $700 billion budget to help it pursue autonomous vehicle technology and is also working on unmanned tanks and smarter vehicles to disarm bombs, many of which will be remote-controlled. Two examples are the Ironclad, shown in the photo above, and Armed Robotic Combat Vehicles made by military contractor BAE Systems, which are expected to have roles in reconnaissance, evacuating injured personnel and disposing explosive ordinance.
Potential pitfalls loom for the military. It will have to navigate differing regulatory structures in foreign countries for things like safety, cybersecurity and liability. It's also encountering resistance from some employee in Silicon Valley who object to sharing their work in artificial intelligence with the military for things like autonomous weaponry.
On the civilian side, automakers are racing to develop self-driving cars even as federal lawmakers continue to wrestle with updating regulations governing them. GM has said it wants to begin mass-producing its self-driving Cruise AV in 2019. But the recent fatality of a bicyclist involving a self-driving Uber test vehicle, as well as fatal crashes involving drivers using Tesla's Autopilot semi-autonomous technology, have illustrated the gravity of the human stakes involved.