Not quite ready for prime time – but already on the roads. From The Guardian:
“Truck platooning” involves two or three trucks that autonomously drive in convoy and are connected via wireless, with the leading truck determining route and speed.
Wednesday’s arrival concluded the first-ever cross-border experiment of its kind, with self-driving trucks leaving factories from as far away as Sweden and southern Germany.
Wonder if it was on highways forbidden to cyclists even before this pilot effort. Driverless trucks would be better on its own separated track from all other vehicles driven and inhabited by human beings.
Love to hear from Dutch cyclists…and car drivers.
I was chatting up with a geospatial specialist at work about this last week ..he shook his head. He felt Google’s contract on liability for their driverless cars, would be thick to cover their butt from damages. And this informal opinion is from a guy with sophisticated knowledge than most readers here on latest in geospatial technology.
This is where overtrust in technology…in the end, is still dependent on human limits in software programming and at arm’s length work. Great for enclosed, smaller protected spaces.
When we can one day guarantee drones that don’t crash, don’t knowingly or unknowingly tread in illegal space, then I’ll believe the safety of a moving 1-5 ton driverless vehicle.
The current system which uses human drivers is certainly a dismal failure. I look forward to new car standards which include active collision avoidance systems. Fully autonomous vehicles will be a natural progression. This is probably the only way to achieve a zero fatality traffic system.
Like SkyTrain, indeed this works well in a very controlled environment, but to operate in the real world will be a decade, likely 2-3 away. I have heard in the 1980’s when I studied computer science that speech recognition would absolve us from typing within a few short years. Ha ! Even a simple command to a car or telephone menu today is almost impossible due to nuances, accents and complexities.
Ever tried auto-translation, auto-spellcheck or Siri for anything beyond weather or restaurants ? Even simple addresses cannot be found with Siri. Loads of room for humans remain. Ditto with trucking.
As such, AVs and ATs (automated trucks) will not be mainstream until maybe 2035 or 2040 due to the many MANY legal and regulatory complexities. Here are four situations ( of many more ) where AVs will have limitations and/or complicated legal / regulatory issues that are not easy to fix, and may differ by city actually:
1) theatre ends and a 1000 people stream onto the street and 100 of them order their AV to the curb at 10:45 pm .. But the curb holds only 10 cars. What will the AV do ? Will we have dedicated AV pick up zones in cities ? Free, or for a fee ? Will they queue up ? Can you pay extra to be at the front of the line ? Imagine the same after a soccer or hockey event with 20,000+ fans and a thousand AVs.
2) AV enters your parkade below your condo. How to open the gate ? Will all condo boards agree on the same standard ? Some use keys or a pad with numbers. That has to be replaced. As such the first and last 100 m of an AV will cost the most to fully automate !
3) drunken dude, a biker or car door slams into AV. Not the AVs fault. Can it continue driving ? Does it have to send the camera report to someone, and if so, when ? Will AV be towed if damaged or can it continue driving, and if so determined by what algorithm ?
4) Firetruck approaches with sirens blazing. Will AV hear it or get an e-signal to get out of the way in a convoluted chaos where cars might drive into sidewalks temporarily to allow Firetruck to pass ?
At YVR there is now a temporary waiting lot (called a cell phone lot) where a driver can wait until the passenger they are meeting is ready to be picked up in the drive through. We’ve used a similar lot at Calgary airport for years. Seems like a pretty good solution for autonomous vehicles, when they arrive. Disagree that it will take 25 years.
Indeed .. might be more than 25 years .. what will AV do if this waiting lot is full, for example (also see my theatre event analogy above) ?
What I CAN see with these ATs is that one truck is driven by a (wo)man .. and 3 – 4 follow closely behind without a driver .. maybe allowed in 10 – 15 years .. from a safe zone to another safe zone .. say a trucking yard or privately owned space from city A to city B !
The idea of autonomous operation for the bulk of the journey, and human control for parking/docking has been used already with large mining trucks. The trucks queue up at the loading shovel. The shovel operator switches control to “unit 1” and with a joystick, backs the truck in. After loading it, he switches “unit 1” to autonomous operation, and off it goes. One operator can run a shovel, and five or six trucks. The same approach can be used for highway trucks. We don’t need to solve the last mile problem to operate autonomously over the rest of the distance. We just have to stop seeing it as all or nothing.
If the destination parking lot is full for the autonomous vehicle? Same as if the roadway is full, from the updated traffic condition channel. Don’t go there. Like a plane that can’t take off because they don’t have a landing slot at the other end. Pretty straightforward. It isn’t like we can keep building more roads for single vehicles. It is bad enough that are roads are full of single occupant vehicles; zero occupant vehicles should be at the bottom of the priority.
Easier said than done in public spaces where all sorts of (human) things happen, unlike a closed zone/private oilsands field, or quarry ..
yes that AV mode will happen gradually like on highways .. but the idea of a completely human less vehicle today happens only in controlled zones (see SkyTrain or CanadaLine) or with remote control (aka drones). So what is probably happening first is remote controlled trucks with no driver .. I can see that in the 2020’s indeed ..
I’m with you on this one, Thomas. Driverless cars clearly work well in a few very controlled environments, academic papers, and popular imagination. The real world is messy and fraught with stupid, dirty hassles. Bringing this technology into common everyday use is the sum of getting thousands of small things right. Twenty-thirty years seems most likely until all this piddling-but-relevant stuff is worked out to the public’s satisfaction. Until then it’s boutique technology.
One way of visualizing the adoption of autonomous vehicles is as an all or nothing, where we go from today (with drivers) to fully autonomous vehicles with no human on board, circling the block. That could take a long time. Several decades represents one view.
Another way of seeing the adoption of autonomous vehicles is as more of a migration to ever more autonomous features. Start with self parking, active speed control, and collision avoidance systems. Reduce road deaths with this technology. Move to autonomous operation on specific sections (highways, dedicated lanes, or zones which have more detailed maps available than the ROW) whereby a driver can engage the autonomous mode, and still be responsible for parking at the end of the trip. Add in platooning, whereby one driver can take five or more heavy trucks on a longer trip, and then park them one at a time at the destination. Add in buses and taxis working on specific routes.
I think there will be a rapid adoption of the second group of autonomous vehicle features. It almost doesn’t matter when the first model is fully adopted, we probably won’t notice it happening.
Since no license is required because you aren’t really driving then you will be able to have that third and fourth, even fifth glass of wine before climbing aboard your AV. On a nice warm night like tonight you’ll want the model with the big sunroof, so you stand up sticking out the top. Your friends will too, so the party can continue as you cruise around town in tandem. In fact there’s no age limit either, nor any need to pass any pesky driving test. Children and grannies too will just clamber into their AVs. Eight year old prodigies will be hacking into the high-end models.
Popular at weed events like yesterday’s English Bay pot party too ..
Sign me up, hombre !
(btw: I prefer beer over wine ..)
If it does in fact become established as a common means of private transport then there will be less, far less, need for rail transit. The Volvo trucks obviously don’t have interest in rail. Substantial funds are being deployed for auto-trucks on roads.
Perhaps the future will look to rail as too limited in its reach, particularly since rain completely fails in the ‘last mile’, whereas AVs can go door to door.
Just how far advanced and just how practical is this technology? Should we put a hold on perhaps wasting money on more subways and rail lines?
Wait, you think autonomous operation is the opposite of rail? Remember Skytrain?
Have to think multimodal, and not get trapped in a box of just one mode.
This idea may work on long, flat runs, say Vancouver to the I-5, then further south. However, I know a couple of long distance truckers who have said the steeper mountainous runs east to the Prairies are very hard on trucks and they (or their freight contract company) charge steeper rates to cover the maintenance.
Trains will always be cheaper because they can carry a lot more weight on rails than a road, the gradients are an easier max of 6%, and their are no bumpy paving repair patches. I had occasion to ride in the front of a freight train from Calgary to Field BC in the 70s — fascinating! Especially the spiral tunnels where a train loses several hundred metres elevation over 14 km partly through a figure eight tunnel blasted through two mountains. In both twists you can see the back of the train entering the mountain directly above the locomotive while leaving the exit portal. Back then the train had over 70 cars and an engineer and brakeman up front, and a conductor back in the caboose. The caboose has been eliminated today and often only an engineer and brakeman operate trains of over 100 cars.
Trucks have a long ways to go to catch up to that.