Cruise Automation's contribution to the self-driving-future:
Started in 2013 and closed in 2016, Cruise automation was never a famous company in the automotive industry, and it will probably never be. But it will go in the history books as the first American company to sell self-driving car technology. From 2015 on their RP-1 kit could convert any Audi A4 or S4 into a self-driving car. A year later the company was bought by General Motors and also the deal set a new record: the auto industry’s biggest Silicon Valley acquisition ever.
Before the RP-1 launched, Cruise barely counted ten employees. But pointing out two of those ten names shows the internal quality they had walking around. First one being founder and CEO Kyle Vogt. The MIT graduate (who of course entered DARPA as a student) was interested in robotics and was described as a genuine hardware hacker. Already as a child he built a 200-pound BattleBot for Comedy Central’s bygone robot-combat competition. Secondly but not less famous is co-founder Justin Kan. Both guys were barely 30 years old when they built both Cruise Automation and video-streaming platform ‘Twitch’ from scratch, which they both sold for about 1 billion dollars.
Their original website http://www.getcruise.com/ has been completely swiped and been turned into a recruitment platform. Before it was showing the first ever American self-driving car set Cruise Automation built, the RP-1. On site you could even pre-order one. The RP-1 could convert any Audi A4 or S4 to a self-driving car, for exactly $10,000. Except for a handful of very early adapters the technology never really made it to the market. And because of the secrecy around the acquisition by GM, not that much is known about the actual technology.
Except that it worked. And that the radar that Cruise used to detect other cars cost around $100 - $200 (the technology is similar to the Doppler radar news channels use to bounce a signal off cloud formations and detect weather patterns). Compare that to $70,000 for Google’s Lidar or even to the $1,000 that Mobileye is spreading around for bringing the cheapest camera possible.
Acquisition by GM
Vogt’s plan was to launch the RP-1 on the global market and to eventually make a kit that would work with any vehicle. A Cruise Automation self-driving car was never the plan. But before he got the time to make that happen General Motors came up with an offer he just could not refuse: $1.000.000.000 dollars! At the time of sale Cruise Automation still only counted around 40 employees and was last valued at around €90,000 post-money. Of course Cruise was the first company in the US to launch a driverless end-product on the market, but that was 10 times the company’s worth and GM obviously acquired Cruise Automation for their technology and not for their end-product. Everyone in the industry smelled something fishy but GM succeeded to keep the secrets well-hidden.
Reilly Brennan’s, leading Stanford University’s advanced auto program called ‘Revs’, theory on GM’s acquisition is that Cruise Automation had deep expertise in ‘computer vision’ and may have found a ground breaking way to analyse and use the huge volume of sensor information coming into an self-driving car. That expertise is hard to obtain and the number of researchers in this area is quite limited. With this data the company could easily have built its own Cruise Automation self-driving car.
In Cruise’s footsteps
The name Cruise Automation will be slowly forgotten while Vogt is working at GM testing the new 'driverless' Chevrolet Bolt EV. But the name will go on forever in the history books. Already it has given inspiration and hope to other startups in the tech/car industry, such as Zoox Inc. and nuTonomy. Both have gained more recognition in the wake of the GM purchase. Zoox Inc. focuses on building a robot taxis with a futuristic design; nuTonomy tells autonomous cars how to plan their motion. Less known examples are Nauto, which is building automotive networking and safety features; or Drive.ai, which is developing artificial intelligence to control autonomous cars.