Cruise Struggles But Progresses Toward Autonomous Chevy Bolt EVs

Cruise Automation Chevy Bolt

OCT 31 2018 BY WADE MALONE 42

Autonomous Chevy Bolt EVs likely to reach many major cities in “single digit years, not decades”

GM CEO Mary Barra recently spoke with The Atlantic at City Lab Detroit about the future of autonomous and electric vehicles. There she again made a call for a national ZEV program. She also re-iterated the need for government – industry partnerships to maintain America’s EV leadership.

The electric transition is inevitable and current Chevrolet models like the Chevy Volt and  Chevy Bolt are just the first steps. Barra says as infrastructure improves and automakers achieve profitability, “Customers are very rational. And they’ll make the right decision. So we need to move on that path. ”

In order to prepare for this, a culture change needed to occur at GM. This has happened over the course of years as they rid themselves of outdated thinking. “There were individuals within the company that challenged – and this was more than a decade ago – the science of global warming. But clearly with the expertise we have in the company, when I’ve been on the leadership staff in the CEO role, we’ve never questioned. So we knew we were on the path towards electric vehicles.”

Barra says their recently announced partnership with Honda for a purpose built autonomous platform is where things get exciting for Cruise. “When you don’t need to have a steering wheel and pedals you can really change the way people move. Make it a more productive space for them.”

That is not to say a steering wheel free model will be the first to hit the streets. Or that steering wheels will go away anytime soon. Future local and federal regulatory requirements are an open question. “Right now, motor vehicle safety standards require a steering wheel, brake and accelerator pedal.”

Cruise Autonomous Chevy Bolt

Like other automakers, GM is still struggling to tame full self driving

Mary Barra says her company already has the ability to scale to mass production of the Autonomous Chevy Bolt EVs. However automakers can not take the immense technological challenge lightly. Their own high safety standards ensure that they will not launch prematurely.

In fact, some former GM and Cruise Automation employees told Reuters that things aren’t progressing as quickly as the company would hope. Primarily in the difficulty of identifying whether objects are in motion or stationary. The software sometimes fails to recognize pedestrians, sees phantom bicycles, and will subsequently brake suddenly.

These are similar issues to those seen on other partial and fully autonomous systems. One GM source claims that “Nothing is on schedule,” including mileage targets and tech milestones.

GM President Dan Ammann said in an interview with Reuters that this is the “engineering challenge of our generation.” But he believes the company is still ahead of the competition. “Right now we are in a race to the starting line,” Said Ammann. “Getting stuck on one particular parameter, or one particular scenario, is missing the fundamental point of what is the total overall performance of the system.”

Cruise CEO Kyle Vogt knows the company is making progress and is in it for the long haul. A market ready fully autonomous service will not happen overnight. But he told Reuters that they’re still on track to hit their 2019 goal.

“With 10 engineers, you can bolt a bunch of sensors onto a car and put a computer in it and get it to drive around the block,” Vogt told Reuters. A commercial product is “about 10,000 times harder.”

Interested in the full talk? Check out the Facebook Live video below!

Source: Facebook Live, Autoblog

Categories: Chevrolet

Tags: , , ,

Leave a Reply

42 Comments on "Cruise Struggles But Progresses Toward Autonomous Chevy Bolt EVs"

newest oldest most voted
ModernMarvelFan

“There were individuals within the company that challenged – and this was more than a decade ago – the science of global warming”

Bob Lutz, anyone?

I guess that was one good riddance….

Dav8or
You know, i have been a Lutz fan for years. He was once a maverick type exec in the auto world. He helped bring good ideas and needed change where he worked but… in recent years he has embarrassed himself by making bold declarations and claims of Tesla’s demise over and over again. It’s to the point that I really think he has been paid to say the things he says on national TV. It’s sad that he felt he needed to stoop that low. I really admire the guy’s work in the past and he still has good insight on the automotive world. As long as the subject isn’t Tesla and Elon Musk, he is actually pretty accurate and informative, but start talking about Tesla and it all goes to some paid anti Tesla spiel. Really sad. And yes, even though he is the father of the Volt program, GM is better off without him now. Mary really seems the right person to run the place and she seems to be building a team to support the move to the future. Of course Bob was never CEO, but he was pretty influential in the day. I thank him for… Read more »
Miggy

GM does not even have an EV (Bolt or Volt) in the World Top 20
http://ev-sales.blogspot.com/search/label/World

dimitrij

That is true, but the article is not about the EV sales figures, but about autonomous driving. Whether they make 2,500 Bolts a month or 3,000 wouldn’t affect the ongoing Cruise R&D effort.

Cypress

They might if you took out all the plugin hybrids from that list.

ModernMarvelFan

“In order to prepare for this, a culture change needed to occur at GM. This has happened over the course of years as they rid themselves of outdated thinking. ”

First step in solving the problem is to recognize the problem. That is why there is actually some hope at GM. But turning around a ship that large such as GM will be hard and takes time.

Keep up the good work, Barra!

Dav8or

No kidding, right? I’ve been a fan of Detroit and US cars for decades. It’s been painful to watch them be brilliant but fail at the same time. A lot of wrong thinking inspired by Wall Street’s short term demands and all parties involved draining the cash cow dry has led to a steady decline of a once American icon to a shadow of it’s former self. Mary seems to be kicking ass and plotting the course for the 21st century. I am thrilled with her tenure so far. I think she really gets it.

Indeed, keep up the good work Mary!

Cypress

“The software sometimes fails to recognize pedestrians, sees phantom bicycles, and will subsequently brake suddenly.”

Also this:

Humans sometimes fail to recognize pedestrians, sees phantom bicycles, and will subsequently brake suddenly”.

DAVID

Ship it.It sounds perfect.

MoMac

Since the article says that the comment (that you quoted) came from FORMER employees — and it seems unlikely that former employees were presenting at the Facebook Live presentation — my rhetorical question is WHEN did these former employees say that?

In other words, is this new information or old information, and how old?

Or is this site really saying that FORMER employees were part of the Facebook Live presentation?

ffbj

Tesla is recognized ev leader in the industry worldwide, so what do need GM for to maintain leadership.
GM talks a lot but doesn’t do much. The Bolt is already old hat, and the Volt is ok, for like 3 years ago, but what have they done lately aside from talk about what they are going to do. Oh we could do this or that or the other thing, sounds a lot like VW and their specious claims. What should we do GM, just exactly what do you want? To improve at 1%/per year mpg, and that’s what’s best for the U.S.A. It used to be, once upon a time where GM could say jump and governments would say how high, but no more.
Put up or shut up as the saying, goes, but I don’t see GM doing much of either.
https://www.youtube.com/watch?v=Kj65AcbekIE

David

Agreed. I’m surprised they haven’t pushed the hybrid power train of the volt into their larger vehicles.

Cypress

They should definitely do that.

ffbj

A number of people have been saying that for years, and I don’t think they are wrong.
Bottom line GM is in the business of making money, and they don’t care how.
Evs don’t make them money so they don’t promote them, but they still need them for ZEV credits.
They would rather sell you an expensive truck, that makes them 10k, rather than a Bolt which loses money, or a passenger car that makes them a little.

MoMac

Don’t conflate a gross loss with a net loss.

When UBS hired Munroe & Associates and released a report on the Bolt, it was clear that Chevy is only losing money on the Bolt when the development and testing cost is taken into account and prorated over cars sold (or an estimate of cars sold in a given period).

As an engineer, when a manager or director (managerial director), tells us that such and such product is losing X dollars per unit, they are always referring to all costs, including development cost.

The upfront development and testing cost is a one-time cost and factored into the cost per car (based on estimated units sold over a period)

The UBS report said the estimated loss per car would be less if the Bolt was scaled up (produced more); which of course means, Chevy has an incentive to sell more.

Cypress

GM has not been idle. You just need to pay attention.

GM partners with schools and aftermarket company to produce an all-electric Camaro:
https://insideevs.com/gm-concept-electric-camaro-sports-car/

Butts in seats program:
https://insideevs.com/chevy-bolt-test-drive-program-geisinger/

Bolt sales up (best selling non-Tesla EV in the US):
https://insideevs.com/global-chevy-bolt-sales-up-nearly-20-u-s-allocation-rises-south-korea-falls/

Producing and selling the Bajoun E100 and E200 in China:
https://insideevs.com/general-motors-baojun-e200-launches-china/

Helping develop ultra-fast 400kw charging:
https://insideevs.com/gm-vehicle-extreme-fast-charging/

Building out charging infrastructure for their Maven fleet of Bolt EVs:
https://insideevs.com/gm-builds-18-stall-ccs-charging-station-for-autonomous-bolt-ev/

Alaa Sadek
Even if GM has enough batteries and at a competitive price, there is an ever bigger problem. Where will they get the software boys and girls? This is a computer on wheels. making a door etc and putting them together to make a shell of a car can be done by robots. Hardly needs people. Not the case with software. AI or no AI. This is not a problem for GM alone? The whole lot of the all over the world. Microsoft Apple, telecom companies etc are all starved. They all need these software guys. We all know that India sells this and did for some time. The software guys sits in Bangalore, writes a program for silicon valley. Are these guys stupid? No. They will want good pay. And Trump is making it even harder for these talents to just enter the US for a vacation let alone work. Does anyone here know of a software company in Germany apart from SAP? Will VW be able to write software? They are good at cheating, we all know that. Will they be able to cheat in writing software too? The same goes for the lot of them all over the… Read more »
GuyMan

It’s called computer science – Lots of schools across the world teach it. The US and Germany have plenty of software engineers. (FANG stocks being a prime example, those are all US companies).

Cypress

Siemens and Bosche are software companies in Germany.

Cypress

University of Michigan in Ann Arbor (1 hour from Detroit) has a a very good computer science and electrical engineering program. It also happens to be collocated with the mechanical, aerospace, and automotive engineering schools. Lots of collaboration projects.

MoMac

GM Cruise is in silicon valley.

I think GM has another group in silicon valley as well.

antrik

“Primarily in the difficulty of identifying whether objects are in motion or stationary. The software sometimes fails to recognize pedestrians, sees phantom bicycles, and will subsequently brake suddenly.”

Wait, wasn’t LIDAR supposed to avoid these kinds of problems, that Tesla’s passive vision approach struggles with?… 😛

GuyMan

No, the sensor isn’t the issue here. LIDAR makes things easier with a point cloud, but cameras can ultimate generate the same data with triangulation (it’s just a harder problem).

The problem is stationary objects, on the road, near the road, besides the road (signs). It’s actually a recognition problem for people as well (lots of rear end accidents), so it’s not suprising that this is a weakness of software.

See: https://www.wired.com/story/tesla-autopilot-why-crash-radar/

MoMac

“identifying whether objects are in motion or stationary”

Reading the Reuters article that the comment was based on, the specific of that problem was that it misidentified and either stopped or slowed down for a parked car that was off to the side (or partially off to the side).

A fault for sure, but not nearly as bad as ramming into stopped firetrucks as Tesla’s autopilot has done 2 or 3 times.

Obviously, this is the exception, not the rule; and the frequency of this occurrence is/was unknown.

“avoid these kinds of problems, that Tesla’s passive vision approach struggles with?…”

So not the same problem as Tesla, at least with regards to “motion or stationary.”

darth

This statement:
“With 10 engineers, you can bolt a bunch of sensors onto a car and put a computer in it and get it to drive around the block,” Vogt told Reuters. A commercial product is “about 10,000 times harder.”

says it all about how far technology has come since the DARPA Grand Challenge days. It has indeed come along way, but getting from a lab prototype to a commercial product is really, really hard.

Richard

Rather than lecturing, GM should get on and build a Gigia factory, roll out a comprehensive fast charging infrastructure and then build a decent BEV with autonomous capability at volume.
After they have done that, then we might take what they have to say on the subject seriously.

dimitrij

Quote: “Rather than lecturing, GM should get on and build a Gigia factory, roll out a comprehensive fast charging infrastructure and then build a decent BEV with autonomous capability at volume. After they have done that, then we might take what they have to say on the subject seriously.”

If EV adoption by general public is the goal, there should be national charging infrastructure, available to all EV’s and not specific to any given brand. Like gas stations and roads.

Brian

If a nation isn’t motivated to switch to EVs then it won’t be motivated to keep a national charging station opperatimg. It will work no better than tire inflators at gas stations. The Supercharger system and copycats are absolutely critical.

Cypress

We really don’t need each manufacturer to have its own charging infrastructure. That’s a recipe to kill mass adoption of EVs. Charging is already confusing to the general public.

MoMac

Copycats?
Huh?

I haven’t seen any charge point that looks like Tesla’s.
There are no “copycats”?

Cypress

Becuase Wall Street looks so favorably on an established company that is making billions, suddenly losing billions. /s

MarriedMalt

Barra says “Customers are very rational”. I couldn’t agree more, that’s why I bought a Tesla Model 3.

jim stack

If customers are so rational and will make the right decision why do they make the Hummer and other HUGE SUV vehicles that get 10 MPG and Roller over very easily?
QUOTE=The electric transition is inevitable and current Chevrolet models like the Chevy Volt and Chevy Bolt are just the first steps. Barra says as infrastructure improves and automakers achieve profitability, “Customers are very rational. And they’ll make the right decision. So we need to move on that path. ”

MoMac

” customers are so rational and will make the right decision why do they make the Hummer ”

LOL — The Hummer hasn’t been produced in the last 8 years.

ModernMarvelFan

Well, GM lost all the irrational customers, some of them stopped buying Hummers, BMWs, and Lexus and bought Tesla instead.

The rational ones are left buying GM cars..

=)