Taking AV to the skies: Near Earth Autonomy landed a defense partnership for its autonomous aircraft tech

Pittsburgh-based autonomous vehicle companies have started to hit the road, but another local firm is hoping to bring the same tech to the air.

Near Earth Autonomy, an autonomous technology company focused on aircraft, announced today that it would launch a new partnership with defense tech contractor L3Harris Technologies to deploy autonomous drone systems for medical response in battle. Founded as a spinout from Carnegie Mellon University at the end of 2012, the company now has around 90 employees and has its local offices in Homewood.

Similar to other autonomous vehicle companies in Pittsburgh, Near Earth develops an autonomous platform that can be integrated into drones or other air vehicles, specifically L3Harris’ FVR-90 hybrid VTOL aircraft in this most recent partnership. With this collaboration, Near Earth said it’s demonstrated a way to deliver blood and other medical essentials autonomously over hundreds of miles. The joint effort to develop this technology was supported through contracts with the U.S. Army Medical Research and Development Command and the Telemedicine and Advanced Technology Research Center.

“The defense department is a good partner at the front end [of technology development] because they are more tolerant to risk,” Near Earth Autonomy CEO Sanjiv Singh told Technical.ly (though that risk tolerance may be relatively new within the federal government). “No one’s going to go bankrupt if a project doesn’t work out.”

That’s important for disruptive technology like autonomous vehicles, particularly when the market for them doesn’t really exist yet, he added. Before the vehicles can reach a more general audience, they’ll need to go through a large volume of rigorous testing that a low-risk defense contract can offer.

Sanjiv Singh. (Courtesy photo)

“This market for large aircraft that can fly by themselves is just coming to be,” Singh said. “And the thing is that absent a big market, the people who are willing to go out and make bets in this area is the defense department.” In comparison to the locally based autonomous vehicles companies that have been making headlines recently — note Aurora Innovation’s November debut on the Nasdaq — Singh stressed that autonomous aircraft is still pretty far from general commercial availability.

All autonomous vehicle developers look to address two main cases, he said: normal scenarios, and scenarios where something goes wrong. In the case of road vehicles, developing the autonomous platform to respond to a problem is easier, because it can direct the car to stop or pull over. The normal scenario, in which a car would need to identify whether something is a bike or a pedestrian or a tree, is much harder to develop.

But for aircraft, “it’s exactly the opposite,” Singh said. Because air travel is already extremely controlled, the general and normal scenarios an autonomous aircraft may come across are relatively simple to account for. The cases in which something goes wrong, however, present a huge challenge. Aircrafts, unlike cars, cannot simply pull over or land in the event of an error or system breakdown.

Because of that, an exact estimate of when autonomous aircrafts will reach the milestones seen in the autonomous road vehicle industry remains unclear. But partnerships like this one with L3Harris are a start. As Near Earth continues to develop its platform for the drones, Singh said it will look to account for and improve use cases for transportation of medical essentials from shore to shore, ship to shore and shore to ship. Near Earth is particularly interested in the results of the latter, he added, as it involves a more complex delivery to a moving target.

Despite these marks of progress, the pathway to general use is long and expensive for autonomous aircrafts. The price point of individual vehicles, paired with the slower progress toward making planes and other aircraft electric, have put it behind the road vehicle industry. But until then, continued tests are the way to prove the feasibility of this tech, Singh said.

In the long term though, “this is how we get to simpler and more reliable and cheaper,” he said. “But that’s going to take some time.”