It was not a matter of if, but when– did anyone think that driverless technology’s first commercial delivery would be a load of 2,000 cases of Budweiser beer?
Scot Bathgate sends in this article from the Toronto Star:
“On Tuesday, Otto, the Uber-owned self-driving vehicle operation, announced the completion of its first commercial delivery, having delivered its beer load from Fort Collins, Colorado, to Colorado Springs, a roughly 190-kilometre trip on Interstate 25.”
It was last August that Uber bought “Otto”, a San Francisco start-up with many employees who had previously worked on Google’s driverless car technology.
“Though largely symbolic, the beer delivery marks the first commercial partnership for Otto, which was founded less than a year ago. Terms of the deal between Otto and Anheuser-Busch InBev, which owns the Budweiser brand, were not disclosed.”
“For this initial delivery, Otto’s truck departed Anheuser-Busch’s facility in Loveland, Colorado, in the early morning before reaching the interstate in Fort Collins. The truck drove through Denver — alongside regular passenger car traffic — and navigated to its destination in Colorado Springs without incident. Otto said a trained driver was in the cabin of the truck at all times to monitor the vehicle’s progress and take over if necessary. At no point was the driver required to intervene, the company said.”
Uber is now going to be testing Otto driverless commercial delivery technology on different road types and weather challenges. The company perceives this driverless technology as a potential game changer for commercial service delivery, with annual trucking industry revenue at $720 billion in 2015. Top brands such as Budweiser owners Anheuser-Busch deliver a million truck loads of beer every year.
No word yet on the future of the Budweiser Clydesdale horse driving team.
This is so awesome.
Before the robots completely take over, there’s going to be this intermediate mode where there’s still a driver in the cab (of the truck, bus, taxi) with something like an “autopilot” (Due to regulations as much as technology limitations). That’s probably going to make driving a much more pleasant job while it lasts…
According to other reports, there was a human in the truck for this delivery, but not behind the wheel. He was reported and shown in photos to be in the sleeper cab.
Likely not. Semi autonomous is proving to be a bad idea. If someone is just sitting there with nothing to do most of the time, they will be reading, on their phone, etc. meaning they will not likely be able to respond effectively in an emergency.
Exactly. In this order of phasing:
1 – foot off
2 – hands off
3 – eyes off
4 – brains off
All will come incrementally ie in stages, first luxury cars, then ordinary one, 1 – 4
Today we have park assist, lane assist, speed control, distance following, simple autopilot in non-stress situations .. and more and more every year.
Fully autonomous AVs further away than you think due to legal issues, complexities and just pure common sense.
For example: theatre is out and 2000 folks leave the theatre and 1000 call their AV to the front curb to pick them up. Can an AV handle that jam today or even in 2035 ?
Or 200 cars are stuck on a traffic jam on a 4 lane road with busy traffic the other side and a crowded sidewalk. Now an ambulance comes from behind and wants to pass through this 4 block pileup of cars with AV sensors blaring ” cars too close on 4 directions” and also AI software will say “sidewalk too crowded”. Will AVs be able to handle that ?
Will they be able to handle snow and ice if parked outside overnight ?
We will see AV modes like airplane autopilot but airplanes still use pilots to land or start or emergency situation !
A-Trucks will for that reason be even further out !
AVs are more likely to be able to handle all those situations better than the random human behaviour that deals with them now. More likely that over time anywhere with crowded sidewalks will have far fewer cars around too.
It seems to me emergency services are unlikely to depend fully on ground oriented motor vehicles in congested areas. More likely a combination of proximity (denser supply of service and built in emergency infrastructure) bicycle mobility, smaller vehicles, and aerial drones.
Snow and ice? Melt ice off sensors and drive with millions of corrections per second better than the best drivers in winter conditions – and more likely to correctly evaluate risk and just stop when necessary. Less likely to proceed into risky conditions. More likely to have other travel options available.
Wake me when it arrives.
Wakey wakey.. stage 1 and 2 here .. in part .. stage 3 and 4 still further out .. see above
Thomas: “A-Trucks will for that reason be even further out”
Welcome to your stage 3. No hands, no feet, human just along for the ride but not able to take manual control.
Autonomous trucks are not as far out as you keep insisting.
This was the actual truck in the story. On public roads.
http://i349.photobucket.com/albums/q367/jcleigh/Posts/Otto%20Autonomous%20Truck_zpsxyymw1vg.jpg
And the driver position:
http://i349.photobucket.com/albums/q367/jcleigh/Posts/Otto%20Autonomous%20Truck%202_zpsnwrluw15.jpg
I would have thought you would have been a fan, given your promotion of Uber, who own Otto.
Of course in sunny weather it is easy to driver “autonomous”. Just wait for these common enough circumstances though:
a) truck needs to gas up
b) road is closed and detour is via clogged, unmapped or narrow roads
c) snow, fog and ice
d) ambulances on streets
e) traffic jams
f) unloading at bay 17
g) a flat tire
Ever wondered why we still have pilots on airplanes ? They too can be flown remotely or via autopilot 99% of the time.
It is the 1% that mucks things up in the real world.
Don’t believe in all the commercials that it will be here “soon”. Progress is always gradual, as the inventor has to make money for it to be used widely. The cell phone was invented in the 1960s but really was used widely only 40+ years later or 50 with smart phones that really changed how we can live, work and interact commercially and socially. And that is a sub $500 device. 50+ years. With a $250,000+ device it will take far longer as trucks once operational will not be replaced as fast as cell phones.
E-cars too won’t come as fast as many think as it is hard to make money cars with very large batteries that last 400km+. So it is often not so much the technology per se, but the ability to make money for the manufacturer. How many robots do you have in your house ? They were invented, and are continuously improved since WW II.
Still waiting for language translation and speech recognition marketed as “almost here” in the 1980s when I studied computer science & AI. When I tell my car to “phone Dave” sometimes it shows me a list of Daves but more often than not it says “pardon”. Try s.th. even more moderately complex such as “Where is the nearest Boston Pizza” and it is silent almost always.
So, yes, in stages, 1, 2, 3 and 4 .. but fully autonomous from truck yard A to customers X, Y and Z unloading & loading goods then back to truck yard B “soon” ? Hardly !
You don’t need a vehicle driver to fuel up, you need a fueling attendant if you don’t want to automate the process. Remember when we had fueling attendants? One worker fueling all vehicles, not one per vehicle.
Autonomous vehicles aren’t working just on preprogrammed routes, they are “seeing” obstacles and recalculating. At a demo around 2008, we placed large boulders in the path of an autonomous mining truck (240 ton). The truck recognized the obstacles, and navigated through them.
Snow and fog? Much easier to design sensors for these conditions (and dark) than trust human eyesight. Improved safety in adverse conditions was a selling feature on mine sites, which tend to be on mountain tops or in deep pits, and often in inhospitable regions.
Loading and unloading? The demo units I watched were autonomous to the dump zone, and to the loading zone. There, they queued up and a single operator (using radio control) “drove” the trucks for a few minutes. One terminal operator handled ten or so trucks, depending on arrival frequency. Straightforward.
Flat tire? My vehicles know about a flat tire before I do. How many people do you think change their own tires these days? Check out the proliferation of run flat tires.
All of these questions have answers. An open mind and a little searching would take one a long way.
Electric vehicle sales insignificant? How about Norway? 25% of new vehicles sold last year? Check out the figures for the last few months there.
Autonomous trucks are $250k each? But the one in the video is a standard truck. The supplier added all the sensors. They aren’t in the truck business, just the autonomous technology business.
I use speech recognition every day. My TV set box has an excellent track record. Maybe you need a new car phone.
Here is a video for those interested. Watch the driver climb out of the cockpit and move back to the sleeper after engaging the autonomous mode.
https://youtu.be/Qb0Kzb3haK8
Author
Reblogged this on Sandy James Planner.
I’m trying to envision as a cyclist…what should we do? Trust that all our roads have shouilders?
As a cyclists, what do you do now? What’s the difference?
I’m also concerned that in the short term there will be deadly mishaps. But in the longer term I’m much more confident that the roads will be safer. AVs won’t show off to their buddies or drive drink or distracted.
Let’s plan for more than just shoulders, such as protected lanes where vehicle volumes or speeds are high.
Autonomous vehicles can be safer around people on bikes than some drivers. A minimum safe passing clearance becomes a programming choice instead of a negotiation.
One of the issues that came up in testing autonomous cars in the city was encounters with people on fixed wheel bikes, with rides doing track stands while stopped at lights or stop signs. The slight rocking back and forth caused the autonomous vehicle to stop and give way to the person on the bike. This demonstrates that current technology can see bikes just fine.
“Let’s plan for more than just shoulders, such as protected lanes where vehicle volumes or speeds are high.”
I guess we will be building bigger and wider roads in cycledom and motordom for some time yet.
To paraphrase Robert Duval, “I love the smell of asphalt in the morning”.
Since the same width of pavement has a capacity for 3 times as many people on bikes as for people in SOVs, there should be a decrease in pavement as more people cycle and less people drive. The extra space can be converted into patio seating or parklets or wider sidewalks. If the transformation of Commercial Drive into a complete street goes ahead as planned, we can see how well this works in Vancouver.
I too love the smell of fresh asphalt, but there is no need to overdo it.
I am really amazed by how quickly this project has developed. There is an article about this in G&M
http://www.theglobeandmail.com/report-on-business/rob-commentary/executive-insight/firms-rushing-to-the-fast-lane-that-is-artificial-intelligence/article32551753/
In less than a year, Mr. Ron (a Google exec) co-founded a company called Otto, built a system that can drive a truck and installed it on an 18-wheel rig that successfully carried 50,000 cans of Budweiser across 190 kilometres of Colorado highway, the first commercial delivery using a self-driving truck. And then he sold Otto to Uber for a reported $680-million (U.S.) and a share of future profits.
The power of artificial intelligence which is at the heart of the self driving software is outstanding.
And troubling. It’s no wonder Stephen Hawking cited AI as one of the threats to humankind. We’ve already discussed how autonomous vehicles will be programmed to select which humans to kill in event of an accident. How far is a leap to a Terminator-like future where the machines decide that violent and erratic homo sapiens is a threat to the planet’s future?
We don’t need bots to come to that conclusion.
How about we do something about it and they might treat us better than we treat the other creatures we share the planet with?
What do you suggest? A cull based on certain criteria?
Germany has now created an ethics committee for these very questions.
More on this here http://www.itas.kit.edu/english/2016_048.php and here http://www.telecompaper.com/news/ethics-committee-on-driverless-cars-to-start-work-this-month–1161888
Specifically, they already decided that property damage is preferred if a human death can be avoided, and that AI software should not prefer one human (say a baby or an elderly or a female) over another human !
Isaac Asimov came up with a set of 3 fundamental Laws of Robotics on which his series of science fiction books was based:
A robot may not injure a human being or, through inaction, allow a human being to come to harm. A robot must obey orders given it by human beings except where such orders would conflict with the First Law. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.