Apple still hasn’t publicly revealed whether it’s working on an autonomous car, but a company started by former Apple engineers certainly is.
Opening up about its plans, startup Aeva has revealed that it is building a smart sensor that will help self-driving cars make sense of the world. And, from its sleek design to its mix of hardware and software magic, it totally looks like the kind of product Apple would design and make!
The tiny white box that the Mountain View-based Aeva is building is capable of measuring objects in a scene, and their distance and velocity relative to one another. The technology is reportedly capable of functioning in all weather conditions, as well as in the the dark. While this won’t handle the actual self-driving part of autonomous vehicles, it will collect the data that informs self-driving cars as they make decisions.
The startup currently employs around 50 people, and has secured $45 million in funding. It is working with a number of big-name automotive companies to integrate its products. Each Aevis box costs around $200 or $300, and a single vehicle may require up to five.
“Typically you have separate LIDAR, separate camera, and separate motion sensors and fuse them in a central compute box,” Aeva co-founder Soroush Salehian told The Verge. “Our product has access to the lowest levels of data. We can measure pixels on certain objects, like a human limb. We can measure the velocity and motion of a pedestrian or object, and we can predict the future motion of those objects pretty accurately.”
Apple’s Project Titan ambitions
Both company co-founders, Soroush Salehian and Mina Rezk, previously worked on Apple’s “Special Projects” team. This included the iPhone, the first Apple Watch, and — quite possibly — Apple’s Project Titan car project.
Apple currently has around 70 self-driving cars on the road, having added 15 new vehicles since May. Despite the project having previously stalled, it seems to now be back up and running again. Current predictions suggest that Apple will launch a so-called Apple Car somewhere around 2023 to 2025.
At the end of 2017, Apple published a machine learning paper, detailing how artificial neural networks could be used to help improve accuracy in LIDAR technology.