California's draft driverless regulations have correctly been described as a step backward for those already testing their vehicles there and for those, like Ford, unpacking their bags in the golden state. Here are some of the gorey details.
How progressive is California if its draft regulations mention ... ?
Use of the term "driver's seat." Isn't the whole point of driverless NOT to have a driver, which would mean there would be no need for a designated seat.
No driverless big trucks and no self-driving interstate buses and no autonomous vehicle that is designed without human driver equipment, such as, presumably, a steering wheel, brakes, and a gas pedal. No futurama designs on California roads.
The devil is also in other details. The regulation leaves open-ended the amount that a manufacturer will be responsible to pay to the state of California for review of an application to go to post-testing use of driverless vehicles, otherwise known as normal use. Really? Does California think these uncertainties will attract driverless research, testing, operations, and manufacturing?
Questionable drivers are welcome, but let's strangle driverless by preventing every conceivable bad outcome
Since licensed drivers are currently free to drive no matter how impaired mentally or physically (unless there's a breathalyzer or other testing device before the ignition can be started), it is interesting that driverless vehicles are required to be equipped with fail-safe mechanisms in the unlikely event of a cyber-attack. So, to summarize: Leave the gun, take the cannoli - okay. Lone brilliant hacker madman who graduated with advanced software manipulation skills - not okay. The first we can litigate in criminal and civil courts, no matter how many people are killed (remember all the times drivers have accidentally or maliciously driven onto a sidewalk full of people); the second is a reason to keep a whole category of vehicles off the road.
Who's in charge?
What will be the third-party testing organizations? These will be the gatekeepers to any driverless activity beyond testing. They are set to make some money IF driverless manufacturers do not first decide to flee the state for more business-friendly jurisdictions. That is a big if. Right now, the qualifications to be a third-party testing organization read like objectively fair qualifications that the author penned with an actual organization in mind. That's my snarky side coming out.
Lots of record-keeping required, which is probably not an onerous obligation as long as the data can be submitted without being completely reconfigured. Maybe it's obvious here that I am not a data person.
Driverless access as a civil right? Not yet in California
The regulations post a discriminatory definition of an "operator" of a driverless motor vehicle. The definition is limited to those physically and mentally capable of obtaining and keeping a drivers license. And the current draft regulations definitely exclude people with visual impairments.There's also a subtext of access to DMV facilities that are difficult or impossible to access by transit. So forget those too poor to get to a DMV.
Good ideas
There are some good ideas in the draft regulations.
1. A motor vehicle equivalent of flight data recorder must be on board every driverless vehicle - good way to reduce litigation, enable early settlements, collect data, and accomplish evidence-based safety improvements.
2. All driverless vehicles should be registered as such.
3. A driverless vehicle may be sold to a museum without a permit, thus ensuring that next to the Julia Child kitchen in the Smithsonian, we will all be able to visit a very early driverless car as well.
4. No beyond-testing use of a driverless vehicle without sufficient insurance coverage for accidents, including serious ones. That goes in the "does that have to be spelled out?" category. Presumably, adequate liability coverage rules apply to every vehicle even the ones we have now. But no harm done in including this.
Monday, January 4, 2016
Monday, December 21, 2015
You Don't Have to Buy a New Car to Be Driverless
One entrepreneur is bypassing the manufacture of a vehicle or the redesign of a current model. Oh no. This idea is basically driverless in a box. His secret - low-cost technology. Connect the sensor box and, presto-chango, your old car is now a driverless car. For only $10,000. When it becomes available. Oh and right now the technology is only designed to work on particular Audi models.
The entrepreneur is Kyle Vogt, the founder of Cruise.
NOT a self-driving car. Cruise only works on the highway. And Vogt plans to expand beyond a couple of Audi models. More information at getcruise.com.
The question is whether driverless-in-a-box can be improved so that it will enable vehicles to be completely self driving.
The entrepreneur is Kyle Vogt, the founder of Cruise.
NOT a self-driving car. Cruise only works on the highway. And Vogt plans to expand beyond a couple of Audi models. More information at getcruise.com.
The question is whether driverless-in-a-box can be improved so that it will enable vehicles to be completely self driving.
Friday, December 18, 2015
On the Street AND Learning, the Vehicle, That Is
I'm a few days late on this, but you have to love the guy who comes up with basically a driverless car in a box that can transform a regular car AND that learns from how people actually drive. This is a different approach than most driverless developers because George Hotz is "teaching the car" via videos of actual driving instead of programming the vehicle with a long list of rules embedded in the software.
Meanwhile, ordering driverless transit
The Swiss public transit driverless pilot in Scion will be starting - fare free - for the general public in the spring of 2016, but some journalists are welcome on board now. Ultimate plans are for on-demand transit that is app-based. According to that post from the Swiss Info Channel, BestMile, the manufacturer, has several cities as potential clients wanting their own pilot driverless transit buses.
Meanwhile, ordering driverless transit
The Swiss public transit driverless pilot in Scion will be starting - fare free - for the general public in the spring of 2016, but some journalists are welcome on board now. Ultimate plans are for on-demand transit that is app-based. According to that post from the Swiss Info Channel, BestMile, the manufacturer, has several cities as potential clients wanting their own pilot driverless transit buses.
Thursday, December 17, 2015
Singapore and England Reveal Ambitious Plans, but not California
Asia
Singapore has an ambitious, organized driverless plan that includes trials for trucks and transit as well as mobility on demand - otherwise known as taxi, ride hailing, or Uber-like service. (Yes, I am starting to use "Uber" as a generic noun.)
One more piece of news from Asia is activity in China. Baidu is set to produce driverless buses to be ready for operation in 2018.
Europe
UK transport officials have met with Google staff multiple times about regulation that will allow driverless transportation on UK roads. Supposedly, England is forging ahead and intends to be a leader. I would add that the Telegraph is the best provider of journalism on the Google/UK topic.
And here is a boring Swiss video from Rinspeed with a very conventional looking car, though it has a retractable steering wheel. Also, it's a personal vehicle, complete with creepy "I'm helping you" - you, specifically - technology. Again, boring. I dare you to make it to the end of the video. I did not.
Step back in California?
California's proposed - not adopted yet - law would require a licensed human driver to be present in every driverless vehicle being operated. Do they not get the whole point of driverless transportation from calling a vehicle to pick up a friend at the station, to deliver pizza, or to park itself? And what about sending the driverless transit bus to the start of its route? I could go on.
If this bill passes, Google, Tesla, and others will concentrate their testing and - maybe soon - sales strategy in other states. Texas, Arizona, and Florida, for top-of-my-head examples, would love to welcome more driverless activity.
Tidbits
Ford to begin testing fully driverless cars on California roads. The car company is partnering with Pivotal Software.
Kia to do driverless testing in Nevada.
Singapore has an ambitious, organized driverless plan that includes trials for trucks and transit as well as mobility on demand - otherwise known as taxi, ride hailing, or Uber-like service. (Yes, I am starting to use "Uber" as a generic noun.)
One more piece of news from Asia is activity in China. Baidu is set to produce driverless buses to be ready for operation in 2018.
Europe
UK transport officials have met with Google staff multiple times about regulation that will allow driverless transportation on UK roads. Supposedly, England is forging ahead and intends to be a leader. I would add that the Telegraph is the best provider of journalism on the Google/UK topic.
And here is a boring Swiss video from Rinspeed with a very conventional looking car, though it has a retractable steering wheel. Also, it's a personal vehicle, complete with creepy "I'm helping you" - you, specifically - technology. Again, boring. I dare you to make it to the end of the video. I did not.
Step back in California?
California's proposed - not adopted yet - law would require a licensed human driver to be present in every driverless vehicle being operated. Do they not get the whole point of driverless transportation from calling a vehicle to pick up a friend at the station, to deliver pizza, or to park itself? And what about sending the driverless transit bus to the start of its route? I could go on.
If this bill passes, Google, Tesla, and others will concentrate their testing and - maybe soon - sales strategy in other states. Texas, Arizona, and Florida, for top-of-my-head examples, would love to welcome more driverless activity.
Tidbits
Ford to begin testing fully driverless cars on California roads. The car company is partnering with Pivotal Software.
Kia to do driverless testing in Nevada.
Thursday, December 10, 2015
One Step Forward for Chinese Driverless Endeavor
China's Baidu company is making headway in its quest to put driverless vehicles out on the roads. Its prototype BMW-model driverless car successfully traveled an 18-mile-plus route under mixed road conditions (no details on that). Now this might seem modest by Google car standards, for example, but (1) there seems to be significant money behind the Baidu program, and (2) once the prototype and the mapping, etc., are in place, early success, in my opinion, belongs to the fastest company that actually sells lots of driverless vehicles (assuming safe road trips). China does not have as complicated or as messily democratic a system of laws and regulations, so the process for legal acceptance of this new mode might well be quicker than in the United States or in Europe.
Not incremental or go for broke
Baidu is taking an approach that differs from both the step-by-step efforts of conventional car companies (and Tesla) and the Google (and Uber and Apple?) strategy of creating a completely driverless, commercially viable vehicle. Baidu's third way is to create vehicles that can handle particular driving environments, such as a particular route - for example, a public bus route. Baidu is relying on extremely detailed mapping, including street signs, as well as "seeing" temporary obstacles, including pedestrians.
P.S. I hope there are no Chinese teenagers who sometimes deface or remove street signs that those rational driverless vehicles are created to read as they travel along. Unexpected consequences and all that.
P.P.S. I admit, post-teenage years, having removed one July 4th sign per child, for their college rooms.
Not incremental or go for broke
Baidu is taking an approach that differs from both the step-by-step efforts of conventional car companies (and Tesla) and the Google (and Uber and Apple?) strategy of creating a completely driverless, commercially viable vehicle. Baidu's third way is to create vehicles that can handle particular driving environments, such as a particular route - for example, a public bus route. Baidu is relying on extremely detailed mapping, including street signs, as well as "seeing" temporary obstacles, including pedestrians.
P.S. I hope there are no Chinese teenagers who sometimes deface or remove street signs that those rational driverless vehicles are created to read as they travel along. Unexpected consequences and all that.
P.P.S. I admit, post-teenage years, having removed one July 4th sign per child, for their college rooms.
Friday, December 4, 2015
Ford + Driverless = Transit?
A European executive at Ford predicts that its first round of driverless vehicles will be for transit or transit-like services because individuals will be unlikely to purchase them and because these self-driving vehicles might not be ready for prime time in all locales due to pesky realities of snow, rain, and darkness. Here are these and other interesting comments from the Ford executive.
GM agrees - or seems to
GM is also now predicting - a departure from previous statements - that within five years, driverless cars will be sold on the consumer market. Within 15 years, they could be the majority of vehicles on the road, GM is saying - at least by a spokesperson in Canada. But GM is also telling Canadians that these vehicles will be shared. (Thank you to Uber and Lyft for changing that conversation.)
And why is GM in Canada to share these predictions? There is movement afoot to do more driverless development and research in the country of our neighbors to the north. The GM president and managing director is plugging for that.
GM agrees - or seems to
GM is also now predicting - a departure from previous statements - that within five years, driverless cars will be sold on the consumer market. Within 15 years, they could be the majority of vehicles on the road, GM is saying - at least by a spokesperson in Canada. But GM is also telling Canadians that these vehicles will be shared. (Thank you to Uber and Lyft for changing that conversation.)
And why is GM in Canada to share these predictions? There is movement afoot to do more driverless development and research in the country of our neighbors to the north. The GM president and managing director is plugging for that.
To Unions: Let's Start Job Training Now
Driverless vehicles - including buses, taxis, and trucks - are coming in a few years. This is a fact, an inevitable revolution about to take place. At some point, the vast majority of human drivers will lose their jobs, whether that comes after union negotiations or at the end of a long contract or whether it is the immigrant Uber or taxi driver being giving the shaft at the end of his shift.
Driver unions can help their members. I hope some politician somewhere powerful can assist those drivers without a union or any worker protections.
Just saying no will not work
However, unions merely declaring that driverless won't happen or that human drivers do so much more than drive and therefore should not lose their jobs - these might stall the inevitable, but they will ultimately be losing arguments.
And the first arguments are just being spouted. An article from Vancouver, shows that the Translink drivers' union and Translink itself are the first in North America to be arguing labor rights as a reason to reject driverless transit.
Their energies would be better spent considering and providing job training to the thousands of drivers who will be out of a job.
Remember actual shoe stores, airlines with good service (outside of first class), small hardware and electronics stores and restaurants before cheap no-frills big box stores and fast food took over? Fortunately or unfortunately, cheap service - that means driverless vehicles without wage and benefits drivers - will also win out. I just hope we remember to replace this step on the economic ladder and to protect those who will be mid-career when they lose their livelihoods.
O' Henry in Australia
This is already happening in a Chinese and Japanese-owned mine in Australia. All vehicles working in the mine are driverless. According to an article from down under, these bot trucks do not get bored and make mistakes. They also eliminate accidents of humans and vehicles working side by side or inside of the the vehicles. [Editor's note: I had trouble getting to the article the second time. Seems the publication prefers readers to be subscribers and not those just occasionally interested in reading a local article.]
Sunshine state
Not everyone is thinking of the unionized drivers or their unrepresented brethren (and sisters). Florida, for example, is doing whatever it can to attract the vehicles and the companies working on them, including hosting an annual autonomous vehicle summit, now in its third year. Here's a video of a self-driving Army vehicle. And an article with more summit details. Okay and a second video below.
Driver unions can help their members. I hope some politician somewhere powerful can assist those drivers without a union or any worker protections.
Just saying no will not work
However, unions merely declaring that driverless won't happen or that human drivers do so much more than drive and therefore should not lose their jobs - these might stall the inevitable, but they will ultimately be losing arguments.
And the first arguments are just being spouted. An article from Vancouver, shows that the Translink drivers' union and Translink itself are the first in North America to be arguing labor rights as a reason to reject driverless transit.
Their energies would be better spent considering and providing job training to the thousands of drivers who will be out of a job.
Remember actual shoe stores, airlines with good service (outside of first class), small hardware and electronics stores and restaurants before cheap no-frills big box stores and fast food took over? Fortunately or unfortunately, cheap service - that means driverless vehicles without wage and benefits drivers - will also win out. I just hope we remember to replace this step on the economic ladder and to protect those who will be mid-career when they lose their livelihoods.
O' Henry in Australia
This is already happening in a Chinese and Japanese-owned mine in Australia. All vehicles working in the mine are driverless. According to an article from down under, these bot trucks do not get bored and make mistakes. They also eliminate accidents of humans and vehicles working side by side or inside of the the vehicles. [Editor's note: I had trouble getting to the article the second time. Seems the publication prefers readers to be subscribers and not those just occasionally interested in reading a local article.]
Sunshine state
Not everyone is thinking of the unionized drivers or their unrepresented brethren (and sisters). Florida, for example, is doing whatever it can to attract the vehicles and the companies working on them, including hosting an annual autonomous vehicle summit, now in its third year. Here's a video of a self-driving Army vehicle. And an article with more summit details. Okay and a second video below.
Subscribe to:
Comments (Atom)