GM/Cruise robotaxis planned to launch in some capacity in 2019, may initially be limited to 30 mph

Mary T. Barra, the chief executive of General Motors, said the company is “on track” to roll out a ride-sharing service in 2019 that would rely on autonomous vehicles, a development that would advance the already-heated race to bring a self-driving car to market.

“We’re on track, with our rate of learning, to be able to do that next year,” Ms. Barra said at DealBook’s Playing for the Long Term conference. She added that the company had a strategy to show how its vehicles are safer than human drivers. The vehicles can currently run safely at speeds of up to about 30 miles per hour, and the service will be limited to a small geographical area, Ms. Barra said.

Limiting cars to 30 mph/48 km/h seems like a smart trick to limit the potential for injuries. I’m guessing there will also be safety drivers and/or remote operators, although that isn’t specified in the article.

Interesting tidbit on lidar range:

G.M. is focused on advancements to lidar, which will drive down costs and enable the vehicles to travel at higher speeds, she said.

One of the problems with most of the lidar units currently available — even some of the expensive ones — is that they are either low-range, low-resolution, or both.

At 113 km/h (70 mph), it takes 91.4 metres to physically stop while driving on wet asphalt or concrete, plus 31.4 metres for every second of reaction time. With a reaction time of 1.0 seconds, that’s 123 metres total to stop. So if your lidar’s range is 120 metres, that leaves you with very little margin for error in terms of reaction time.

Much like Waymo’s announcement of launching a commercial service I’m hopeful to see these services becoming real, but not holding my breath. I would expect Waymo and Cruise to have provided much more public demonstrations of the capabilities of their systems if they were actually ready to field a real commercial service. It’s hard not to think that secrecy is a symptom of lack of confidence.

Unlike Tesla’s model, which I’d characterize as rapid physical deployment followed by gradual capability increase, a taxi service’s natural development path is gradual physical deployment starting from a basis of extremely high capability. Since taxi service has to have high capability a priori it’s the capability level itself which limits the start of ramping the service. If a taxi service was ready to start making real impacts on the world it would have to have already solidly resolved its technical capability problem, and there wouldn’t be so much need for secrecy.

I’m imagining that we’ll see these taxi services operating in small whitelisted areas for several years before broader deployment happens, and that Waymo / Cruise are expecting that as well. Given that expectation they should be anxious to get out into public as quickly as possible. Their failure to do so leads me to believe that they are still suffering from a lot of problems with their capability.

1 Like

Waymo is apparently going to release some of its early riders from their NDAs sometime soon, meaning they can vlog their trips or bring members of the media along. That would give the public some real, unfiltered insight into how the system performs.

I’m very much looking forward to learning more.