A big test of the ability of autonomous cars to replace taxis and ride-sharing services is on the way.
VentureBeat reports GM reportedly plans to build and test thousands of self-driving Bolts in 2018.
General Motors plans to deploy thousands of self-driving electric cars in test fleets in partnership with ride-sharing affiliate Lyft Inc, beginning in 2018, two sources familiar with the automaker’s plans said this week.
It is expected to be the largest such test of fully autonomous vehicles by any major automaker before 2020, when several companies have said they plan to begin building and deploying such vehicles in higher volumes. Alphabet Inc’s Waymo subsidiary, in comparison, is currently testing about 60 self-driving prototypes in four states.
Most of the specially equipped versions of the Chevrolet Bolt electric vehicle will be used by San Francisco-based Lyft, which will test them in its ride-sharing fleet in several states, one of the sources said. GM has no immediate plans to sell the Bolt AV to individual customers, according to the source.
The sources spoke only on condition of anonymity because GM has not announced its plans yet.
In a statement on Friday, GM said: “We do not provide specific details on potential future products or technology rollout plans. We have said that our AV technology will appear in an on-demand ridesharing network application sooner than you might think.”
Lyft declined to comment.
Sooner Than You Might Think
The time-frame for “sooner than you might think” has now advanced to 2018.
GM executive Mike Ableson stated “If you assume the cost of these autonomous vehicles, the very early ones, will be six figures, there aren’t very many retail customers that are willing to go out and spend that kind of money. But even at that sort of cost, with a ridesharing platform, you can build a business.”
No doubt the skeptics will point to the cost. But that is the 2018 cost. It will not be the 2020 cost.
The price of technology is dropping like a rock. For now, if you are only going to build a few thousand vehicles, the cost is bound to be high.
Ford announced production for 2020. GM just leap-frogged Ford by two years. Competition is on, and it will get more intense.
The first big payout will be trucks. Millions of long-haul truck driving jobs are in the crosshairs of major disruptive force. Those jobs will vanish “sooner than most think”, no later than 2022 to 2024, and possibly sooner.
Mike “Mish” Shedlock.
It’s coming and I can’t wait until the long haul trucks stay out of the left lane!!
I would rather have one of these:
https://www.pal-v.com/en/
This is going to be the best thing ever – better than the PC and the cell phone! Can’t wait.
Why the continuous need to announce vaporware?
Because banksters spend more of their money on stocks and bonds sold on the back of hype, than they do on actual product. While non banksters no longer have any money to spend on anything, hence are no longer a market worth serving.
Depends on how you define “autonomous”. There are no level-5 autonomous vehicles that can handle every driving situation a human could. Those are decades away. There are some level-4 systems that can handle most driving situations under most conditions but still must hand control over to a human if it encounters a “tough” situation. Level 3 is where most systems are where a driver must still be present at all times and ready to intervene for the autonomous system.
level 4 already here.
If drivers have to handle some situations, so what? The vast majority of uses won’t, and that’s where the jobs are.
If the automation can know when to hand control to a passenger, then it can know to contact an operator at the automation’s call center to look at what the car sees (in 3D) and remotely drive the car back into the automation’s safe “envelope”. It would also know when to call police, fire, ambulance, a mechanic, animal control, et cetera. The remote human may be able to check external traffic cameras (maintained by government — with the vehicle they’re driving automatically identified in the image) and even the onboard automation cameras of other driverless vehicles; including those of another marque or with different software.
The conditions that caused the incident will later be reviewed to identify needed improvements and enhancements in the software, which will make the software a better driver. That’s how humans learn to drive, but while we do it as individuals “the system” will apply the lesson of one to the performance of millions.
Let’s just hope the remote driver isn’t in India, where driving rules are more suggestion than law.
The driverless truck will be a much bigger deal.. With fewer schedule constraints, long haul traffic will be scheduled to pass through cities during low traffic hours.
Big improvement to city traffic.
big improvements in safety
last year more than 40,000 Americans were killed by human drivers
Time honored and proven, the already existing rail system easily solves the long haul problem. Clearing the rails of tanker cars which transport crude oil that belongs in pipelines and you suddenly expand capacity for freight without laying down even another foot of track.
Big improvement in city traffic? Absolutely.
Big improvement in city safety? Depends how you feel about millions of gallons of crude rolling through your streets and byways in tanker cars.
The Bolt is a horrible vehicle, cramped and short range, but it’s what GM has available for testing ….
Luddite fatigue.
Most families have multiple cars and one being all electric now makes sense.
Looking forward to 2022 so I can post “So where are all these self-driving cars, Mish?”…
Meh. There’s a statistically-likely chance you will have been killed by a human driver by 2022.
I am optimistic but also pragmatic.
Please have an article about “what ifs” concerning self-driving cars or trucks.
I start:
What if a tyre puncture ?
What if any mechanical anomaly ?
What if obstacle in front ?
What if very bad weather limiting drive sensors ?
What if hacking ?
What if creating accident ?
What if given order by authorities ?
What if hauled content needs any special watch ?
The general question is what if need to react unexpectedly. No one to take responsibility then, except if supervised one to one remotely.
I believe in self-driving but it would be a complementary system, so dedicated (very) limited roads or self-driving with still a driver in careful watch at least.
Of course 10 mile straight road in desert in Australia or America are fit for that purpose. Nowhere to find in crowded parts of the world.
New concept which will develop bit by bit for sure in long term projects.
Rome was not built in a day (or decade if you prefer)
The level of security against hacking would have to be very good. I assume a truck fleet would be controlled by a home base system and ultimately monitored by humans. Maybe with ability to drive remotely under unusual conditions, Kind of like how predator drones are remotely controlled. All the video game playing will come in handy.
They will be hacked. What is more fun than hacking into the system and driving it off a cliff. What better terrorism than hack the truck and run over a few people.
I suspect we will get our infrastructure $$$ when the owners of autonomous fleets realize that socialized investment will support their private bottom lines. Better late than never I guess.
What good are automated trucks when our roads and bridges are falling apart.
Now it’s a race to the bottom, er, I mean to be first.
Walmart quality & pricing to follow.
For those doubters who cite cost, think back to flat screen TV’s. I remember seeing 50″ plasma TV’s selling for $20,000 in the ealry 2000’s. The price may be high now, but it will come down.