Apple has been working on autonomous vehicle technology for some time now.
Documents from August 2015 revealed that the company was developing some sort of self-driving car tech, while the company was granted a patent in April 2017 to test such vehicles. CEO Tim Cook has also been quite bullish on self-driving cars, calling the tech “the mother of all AI projects” back in June.
That said, the company has been rather secretive over the years when it comes to its self-driving vehicle efforts. To shed some light on the matter, Apple’s director of artificial intelligence research, Canadian Ruslan Salakhutdinov, met with around 200 AI experts in an event this past weekend.
Salakhutdinov referred to a research paper published last month, which detailed software to identify pedestrians and cyclists using 3-D scanners called lidars used on most autonomous vehicles. Salakhutdinov mentioned that thanks to machine learning, the system was able to perform well even during rainy conditions or when pedestrians were partially obscured by parked cars.
Salakhutdinov also discussed a positioning technique known as SLAM, which is used on robots and autonomous vehicle for simultaneous localization and mapping. SLAM also features map building and augmented reality applications.
Finally, Salakhutdinov talked about how Apple is working on using AI to let vehicles make quick decisions in certain situations, such as avoiding a pedestrian on the road. Data collected by cars with large mounted sensors has been used to create 3D maps that feature traffic lights, road markings and other mapping information.