There's a reason there are consumer autonomous systems already being rolled out for highway driving without an equivalent for city driving. Cars on a highway behave rationally for the most part, which makes it easier for the artificial intelligence that controls an autonomous system to operate since it is programmed logically.
Author: Sam Chase, The Connected Car
In a city, the density of vehicles is much higher, which leaves less margin for error than on a highway. There are also far more irrational actors, such as pedestrians and cyclists, whose movements cannot be predicted with great accuracy.
That can force an autonomous system to react much quicker than it is capable of doing currently, and the reaction may not always be the right one. The level of artificial intelligence necessary to operate a driverless car in rush hour city traffic is higher by an order of magnitude compared to highway driving.
For autonomous vehicle developers, this type of AI challenge is to be welcomed, not avoided.
Operating out of San Francisco, General Motors self-driving subsidiary Cruise Automation has aggressively tested its driverless vehicles in complicated urban settings for quite some time. According to a slideshow released on the company's website in advance of an investor presentation in late November, the company is scaling up its efforts in cities by launching a commercial AV service in 'dense urban environments' starting in 2019.
The presentation imagines a future with zero crashes, zero emissions and zero traffic congestion. It is relentlessly optimistic, and positions GM as a leader in autonomy and the 'only fully integrated developer of AVs with true scale ability'.
The reasons GM puts forward for its AV superiority range from big-name hires to impending technological breakthroughs. On the talent front, GM has recruited former Uber engineer AG Gandgadhar as the first ever CTO of Cruise, former Netflix executive Tawni Cranz as its new chief HR officer, and former Google Maps programme manager Ashwin Prabhu as head of mapping operations.
On the tech front, GM says its Lidar developments are game-changing. While currently available Lidar systems are described as having quality issues and a $20,000 price tag, Cruise claims its acquisition of Strobe will enable it to manufacture a Lidar system that is more than twice as effective as current products at a cost of only $300 per unit.
In October, Cruise CEO Kyle Vogt published a Medium post underlining the philosophy that drives his company to pursue deployment in urban environments specifically.
"Our path to scale starts with fleet deployments in the most dense urban environments, where high utilisation rates generate sustainable unit economics even at low initial vehicle volumes," Vogt wrote. "As we ramp up volume, we'll drive down costs and gradually roll out our technology to suburban and rural areas where ride sharing is less common today. It takes scale in order to reach these areas and achieve a significant impact on society, so for us anything less than that is failure (and probably an unsustainable business)."
There was no mention of which cities Cruise would be expanding into. Wherever they end up, those cities will instantly become autonomous hubs, which could lead to a network effect that creates even faster innovation in the category.