What can AVs learn from industrial robotics?

The realisation of a driverless mobility ecosystem has proven harder than originally anticipated. Years ago, researchers demonstrated vehicles that could technically drive without a human at the controls, but the challenge was far from over.

“A lay person may see a ‘self-driving’ car in public and think that’s the problem solved,” observes Roboticist Daniel Theobald. “I’m using air quotes here for a reason—making a vehicle move without someone in the driver’s seat is incredibly easy. The challenge boils down to the 80/20 rule.” This refers to a situation in which a researcher spends the first 20% of their time solving 80% of the problems, and then 80% of their time solving the last 20% of the problems. In self-driving cars, suggests Theobald, that ratio is more like eight million/20.

“Part of the problem is that there’s no solving autonomous driving. It’s going to be a constant evolution of taking on more and getting better at it,” he tells Automotive World. And luckily, there’s an established industry that can offer plenty of lessons.