Apple Inc. (AAPL) has finally begun to open up about its secretive self-driving car technology after years of keeping investors in the dark.
At the Neural Information Processing Systems conference in California, the iPhone maker’s artificial intelligence (AI) research director Ruslan Salakhutdinov revealed that Apple is using machine learning to analyze large stockpiles of data from cameras and other sensors, according to Wired. Thanks to this data, he claimed that the company’s software is now able spot cars, cyclists and pedestrians on busy streets, navigate in unfamiliar spaces and build detailed 3D maps of cities.
Salakhutdinov believes Apple’s onboard cameras are sophisticated enough to identify objects in the trickiest of situations. He said that the cameras are fully functional even when raindrops cover their lenses and that 3D scanners called lidars can estimate the position of pedestrians even if they're hidden from view.
Apple’s AI research director explained that cars are given directions through a technique known as SLAM, simultaneous localization and mapping. He added that sensors contain detailed 3D maps featuring traffic lights and road markings and the ability to make decisions in urgent situations, such as when a pedestrian appears out of nowhere.
Salakhutdinov described these latest developments as Apple’s autonomous driving team’s biggest breakthroughs yet. “If you asked me five years ago, I would be very skeptical of saying ‘Yes you could do that,’” he said.
According to Wired, Salakhutdinov refused to confirm whether the projects he discussed during his presentation fit into any wider effort in automated driving. He also failed to reveal the scale and scope of Apple’s car projects.
Apple’s self-driving projects have been shrouded in secrecy. The company’s CEO Tim Cook confirmed his interest in the technology in June, shortly after Apple received a permit from the California Department of Motor Vehicles to start test driving self-driving vehicles. Little other information has been provided until now. (See also: Apple Shifts Self-Driving Vehicle Plan from Cars to Buses.)
Salakhutdinov's decision to become less secretive about his employer’s self-driving ambitions is believed to motivated by a desire to attract the best AI talent. Researchers in the field are currently said to be more tempted to join the company’s tech rivals, such as Facebook Inc. (FB) and Alphabet Inc.’s Google (GOOGL), as their projects are more out in the open and, as a result, tend to generate greater recognition. (See also: Self-Driving Cars to Benefit These Stocks: Morgan Stanley.)