NASA brainboxes work on algorithms for 'safe' self-flying aircraft

It's the fear of anyone who watches Snakes on a Plane and books a flight – what if your plane crashes? Now take a deep breath and imagine that you're travelling on a plane or rocketship with no pilot. A new NASA research project hopes to find ways to certify unmanned autonomous aircraft systems for safety.

The easy part, says Carnegie Mellon University computer scientist Philip Koopman – who studies the safety of autonomous systems such as cars or drones and is not involved in the research – is when the plane is cruising. One reason self-driving cars have a lot of trouble getting certified for safety is that they often have little time to react to obstacles. (Concrete wall. Boom.)

Although safety will require much more research and planning, and planes can't simply "pull over" like cars, the skies are a lot less crowded than the ground, he says, so planes will naturally have more time to react or use failsafes (such as circling). Take-off and landing would be the hardest to validate for safety because aircraft sometimes only have half a second to react, he says.

"The technology has to mature some," he says.

The NASA feasibility study seeks to create algorithms that can help give users confidence in the decisions they make, in order to some day help autonomous system certification, according to a press release.

It's one of three new investigations this year for NASA's Convergent Aeronautics Solution project, which conducts short-term tech experiments to try to solve problems in aeronautics. Others include making sure drones are ready to fly and quantum communication networks for drones. The autonomous system certification feasibility study will conclude in about 24 to 30 months. ®

*However, before you pshaw at the lack of self-flying planes, there will be some very cool self-flying object collision maths involved, and we'd argue that is very cool too.