The military will have driverless vehicles before the rest of us – or that’s what the US undersecretary of defence for research and engineering claimed in 2019.
More recently, Elon Musk has been saying Tesla will be able to create fully autonomous vehicles within the year.
So it seems the race is on between business and the military to come up with the first truly autonomous vehicle (AV). Here, we take a look at how the military is doing.
The defence sector has traditionally been the leader in AV development and has been experimenting with the notion for decades. But in the mid-2000s, the US Defense Advanced Research Projects Agency (DARPA) ramped up its interest in AVs, hosting a series of races and competitions for driverless vehicles. Universities and businesses took the bait and got involved, leading to breakthroughs of their own.
These days, advances in AV technology in the commercial sector are beginning to flow the other way, back to the military. Previously, sensor technology was prohibitively expensive, but thanks to industry’s efforts to develop the tech for a larger market, the cost is falling for the military.
Despite Musk’s confidence, most experts believe we’re some way off seeing genuine autonomous vehicles on our roads. Regulatory and ethical problems will need to be solved before commercial adoption; however, these issues will land differently in the military, where swapping vehicles full of troops for driverless ones will obviously reduce casualties during reconnaissance and supply missions. In the US, research by RAND corporation suggests that a fully autonomous convoy would put 78 fewer lives at risk.
While the tech for this is not yet available, the US military is testing out leader-follower technology, in which 12 or more AVs follow a two-man vehicle around like giant robot ducklings. This would put 28 fewer troops at risk, says the same study.
The military is also expected to push ahead with remote control, or teleoperation. Remotely controlled vehicles would have obvious advantages in a warzone, where they could carry out surveillance and search and destroy missions for explosive devices. SInce the commercial applications are less obvious, the military may take the lead here.
However, there are bigger engineering challenges for military AVs, such as inhospitable environments and rough terrain, which means extensive testing is needed to programme the A to respond correctly. Military vehicles also need to be able to navigate combat zones where there may be no road markings, signs, or high resolution maps.
Couple this with the lack of 5G infrastructure in war-torn countries, and the likelihood of enemy cyber attacks on military AVs, it seems that fully autonomous military vehicles are still some way off.