GM's Billion Dollar Distraction

The Automotive News informs us that General Motors' Cruise Automation division claims to have created the first true self-driving car. Okay, well, not exactly. The article immediately dives into a variety of caveats from Cruise founder Kyle Vogt who apparently misses the yawning chasm standing between his cobbled-together people carrier and a meaningful means for finding a market.

https://tinyurl.com/y8gsuula - "GM's Cruise Automation Touts Bolt EV as 'First Real Self-driving Car'" - Automotive News

https://tinyurl.com/ycyjogfo - "How We Built the First Self-Driving Car (Really)" - Medium

The car lacks some key elements to be truly self-driving - ie. driverless. Given the findings released yesterday by the National Transportation Safety Board regarding the fatal crash of a Tesla Motors Model S in 2016 one would expect a more circumspect air to pervade the world of autonomous driving. Wishful thinking that.

Vogt is not one to curb his enthusiasm. He goes so far as to write in his Medium post that if Cruise had followed "typical OEM development cycles" it would have taken six years to produce what the company has delivered in less than one. There is one big caveat which is that part of the development cycle of any vehicle is understanding the target market.

Vogt writes that he wants to contribute to keeping GM's Bolt plant busy making lots of Bolts presumably with Cruise's bolted on hardware. But what is the market for this vanity project?

Bolts are now being sold nationwide and are also being activated as part of Maven's car sharing fleet, but what is the point of a self-driving Bolt or, more accurately, a nearly self-driving Bolt? At least Waymo has taken the steps to put out marketing messages regarding use case scenarios featuring families in Phoenix using Waymo vehicles to get around. Prior to Waymo Alphabet produced videos showing older or blind passengers getting around in the predecessor self-driving pods.

But what is the vision for Bolt? The Bolt has limited application as a mover of large numbers of people in its hatchback configuration. It is unclear how much room is left in the car after the installation of the hardware. Is it really ready for a driverless ride-hailing deployment? It doesn't look that way.

In essence, Kyle Vogt is the dog that caught that car he was chasing. Now what?

There seems to be an utter lack of vision. The true irony of deploying the Cruise technology on the Bolt is the fact that the Bolt is actually fun to drive - as are most EVs with instant acceleration and whisper quiet operation. Why would you want it to drive itself?

But the answer lies in a presentation recently shared with me by Uri Levine, founder of Waze. Uri points out that we need to get more people into fewer vehicles if we ever hope to solve the traffic congestion crisis gripping the world.

Uber, Lyft and the rest of the carpool crowd are not the answer because potential passengers are not eager to share their ride in a small vehicle with a stranger. Autonomous vehicles are not the answer, because by the time they become ubiquitous there will be hundreds of empty peopleless cars floating around major cities chasing after the next ride or returning to a charging station.

The solution is larger capacity shuttles either driven or driverless. The Chevy Bolt simply is the wrong car for this task. Vogt's gonzo approach to autonomous tech has been pursued in a vacuum and without the benefit of GM's broader marketing resources and distribution network. Word is Vogt chased off early envoys from GM insisting on his own autonomy.

The Cruise-equipped Bolt's are the product of a mad scientist too expensive (and unattractive) for anyone to purchase and ill-suited to public transportation applications. There is little doubt that much has been learned in the creation of the retro-fitted Bolts. But it is time for Vogt & Co. to sit down with the engineers and marketers at GM to figure out where Cruise fits into GM's evolving vision of autonomy.

Here it is worth noting that GM earned a tip of the hat in the NTSB's report on the crash of the Tesla Model S. The agency found Tesla's system for ensuring a driver was paying attention - detecting hands on the steering wheel - was inadequate. The agency specifically recommends more direct driver monitoring solutions along the lines of GM's recently announced Supercruise technology from Seeing Machines. Now that is progress.

Roger C. Lanctot is Director, Automotive Connected Mobility in the Global Automotive Practice at Strategy Analytics. More details about Strategy Analytics can be found here: https://www.strategyanalytics.com/access-services/automotive#.VuGdXfkrKUk