Driverless Electric Truck Starts Deliveries On Swedish Public Road

Reuters reports that a driverless electric truck has begun daily freight deliveries on a public road in Sweden. Slashdot reader Kiuas shares the report: Robert Falck, the CEO of Swedish start-up Einride, said the company was in partnership talks with major suppliers to help scale production and deliver orders, and the firm did not rule out future tie-ups with large truckmakers. "This public road permit is a major milestone [...] and it is a step to commercializing autonomous technology on roads," the former Volvo executive told Reuters. "Since we're a software and operational first company, a partnership with a manufacturing company is something that we see as a core moving forward," he said, adding he hoped to seal a deal by next year.

Einride's T-Pod is 26 tons when full and does not have a driver cabin, which it estimates reduces road freight operating costs by around 60 percent versus a diesel truck with a driver. The T-Pod is level 4 autonomous, the second highest category, and uses a Nvidia Drive platform to process visual data in real time. An operator, sitting miles away, can supervise and control up to 10 vehicles at once. The T-Pod has permission to make short trips - between a warehouse and a terminal - on a public road in an industrial area in Jonkoping, central Sweden, at up to 5 km/hr, documents from the transport authority show. Falck said Einride would apply next year for more public route permits and was planning to expand in the United States.

Worse than b-----

[phantomfive]

on a public road in an industrial area in Jonkoping, central Sweden, at up to 5 km/h

May I please not get stuck behind that thing.

Re:

[Anonymous Coward]

May I please not get stuck behind that thing.

That's just what horse and buggy drivers said about those gosh darned newfangled "horseless carriages". Too damn slow to be allowed on the roads.

Re:

[phantomfive]

That's just what horse and buggy drivers said about those gosh darned newfangled "horseless carriages". Too damn slow

No one said that. Stop making things up. If this Wikipedia article is accurate [wikipedia.org], it was just the opposite: people complained that the cars were too fast, and set the speed limit to 3.2 km/h.

Re:

[0100010001010011]

You could just pass it? It'd be like coming up on a tractor in farm country. Just go around it.

Re:

[Rick Schumann]

3.1 mph. Literally walking speed, if you walk slow that is. Most people could pass it on foot without having to break into a run.

Re:

[gweihir]

Of course, this is just the initial speed. They had to make it working on a public road, they had to get permission and this is the first large-scale experiment of this type. The vehicles chassis likely support much faster speeds and it is entirely reasonable that they will progressively get faster. This may also require upgrades to the electronics.

Re:

[JaredOfEuropa]

Hopefully it won’t be cancelled after too many unsuspecting drivers rear-end that thing...

Re:

[gweihir]

This will likely be an area with a 30km/h or lower speed-limit anyways.

Re:

[AmiMoJo]

The speed limit is probably due to the range of its sensors.

At that speed if the ultrasonic sensors trigger it can stop safely before hitting something. They are short range but it is going very slowly. They don't trust the camera system or lidar or whatever it has.

Re:

[gweihir]

And they are right to not trust the long-range servers. Yet. They need to collect data on reliability, false alerts, etc. and this deployment allows them to do that. Then they can go to, say, 10km/h and repeat. This way, they will hit regular speeds in something like 10-20 years or find out they need a different approach. That is about normal for engineering progress with new technologies.

Re:

[cascadingstylesheet]

on a public road in an industrial area in Jonkoping, central Sweden, at up to 5 km/h

May I please not get stuck behind that thing.

On my eeeenternational Harverster!

Cussin' at the robot ain't gonna save you no time, hoss!

So tell me again

[rsilvergun]

how the robots aren't coming for driving jobs for another 20+ years. I'm all ears. And I got all the time in the world what with the global economic crash that's coming when 10% of the work force disappears almost overnight...

It's like folks keep saying: If we don't have jobs who'll buy their products? Please. Ask yourself, did the King need peasants to buy his products? No. He was the King. He already owned everything there was to own. He's King.

Re:

[K. S. Kyosuke]

If you're not people-limited, why "hit 70 mph"? As far as non-perishable goods are concerned, you don't necessarily need them moved at highway speeds.

Re:

[Rick Schumann]

You can start worrying when they're on the highways at full highway speeds and travelling coast-to-coast unattended. Before then however I'm sure we'll see some horrific accidents when 40 tons of fully loaded truck screws up royally and causes a massive accidents, likely killing people. I have little faith in this 'technology', I think it's over-hyped and half-assed at best.

Re:So tell me again

[slack_justyb]

Before then however I'm sure we'll see some horrific accidents when 40 tons of fully loaded truck screws up royally and causes a massive accidents

Ugh, this logic is so frustrating because it so heavily relies on feelings and not reality. Reality is that semis cause thousands of deaths around the world every single flipping day, and yet we still use them?! Why? Because paying out on the deaths, the lawsuits, insurance, etc is far cheaper than trying to operate a world without semis. And the exact thing is true here. Let's just assume that automatic semis at first only have 10% less accidents. Well, that's a massive cost savings! So yeah, let's just say self driving semis the first twenty years in kill 800 people every single day in the US alone. That's still a massive positive versus the thousands that die with the non-self driving versions.

Typically when you sit down and crunch numbers, everyone inevitably ends up with some argument of "well something something if you die form them, or I don't want to die from them..." and all those arguments are just "feels over reals" arguments that you just can't have a debate on. Yes, I get it, no one wants to die. However, the reality is that we're killing people left and right here on the highways and the objective is that we'd like to kill fewer. We're not going to ever reach zero deaths, but shit, not killing 10% of the people that Dan the Truck Driver that was taking a peek at Facebook real quick would have mowed down is still better.

So yeah, you can use adjectives like "horrific" to nail in that extra punch of emotion there, but the reality is this.

likely killing people

We already do that in spades. Tons of folks on this planet met their end on the road, so what outrage are you thinking is going to be coming? The only outrage there will be is the manufactured one because a "robot" did it. And then we'll all move on with our lives and the robots will get better. And then more people will die and we'll act all shocked, and then we move on with our lives and the robots get better. And then some more people will die and more after that and more and more and more and more and after every death we'll hear about it on the news a little bit less and the robots will get a little bit better. All the while we'll all just keep moving on with our lives because... We, the public at large, don't honestly give a fuck about the thousands of people who already die every single stinking day currently. That's not even in the top 100 line items on any political agenda. So spare us this little diddy you're playing on your tiny violin there about how horrible it will be when all these people needlessly die.

Re:

[Rick Schumann]

Yeah? I'm goddamned sick and tired of having to write out entire paragraphs explaining the how and why this so-called 'self driving car' technology is half-baked at best, how they're trying to shove it down our throats regardless, and how much it's going to FAIL, SPECTACULARLY, and how people are going to DIE, SENSLESSLY because of it. For fuck's sake stop listening to the marketing and media HYPE and actually go out and do your own research. I'm tired of having to explain it to everyone.

Re:

[gweihir]

Then stop. Because you are wrong. And sure, it is not a finished, mature tech at this time, but it is clearly getting there. If only mature tech saw widespread use, we would not have most tech.

Re:

[Rick Schumann]

Bullshit. It's half-assed, can't think, no capacity for cognition whatsoever, can't tell the difference between a lamppost and a living being, and so on, and so on, and so on. Be sure to enjoy screaming in abject terror when your Level 5 SDC fucks up and drives you at 100mph into a concrete abutment, or runs over a pedestrian and you can't stop it from happening. You people will have to learn the hard way I guess.

Re:

[serviscope_minor]

Bullshit.

No he's absolutely right.

It's half-assed, can't think, no capacity for cognition whatsoever,

At the speed at which driving requires humans to make decisions, neither an they most of the time.

can't tell the difference between a lamppost and a living being, and so on, and so on, and so on.

But you're ignoring that it never gets tired, never gets frustrated, never gets distracted, doesn't answer text messages or check facebook, never yells at the kids, never speeds, i never late for anything, never g

Re:

[gweihir]

I don't have very much faith in my fellow humans and it drops a bit every time I get on the road. I don't think that SDCs will have to be particularly great before they have fewer accidents than human drivers.

And that is just it. Humans are really bad drivers and kill a lot of people all the time. And then they do stupid things like driving intoxicated or playing with their phones while driving. Sure, autonomous driving is hard, but serious research into it has been going on for about 40 years now and it is having results. We can now do level 4 (with _remote_ assistance to the car if needed) at 5km/h. And this will get both faster and eventually it will be at level 5 for all standard situations. May still take 1

Re:

[serviscope_minor]

And that is the killer: If autonomous cars kill 1/10 of the humans than manually driven ones do, that is a huge, huge win. If nobody else, insurance companies will kill the driver-controlled car eventually, or limit it to the 5km/h the autonomous ones currently have to observe.

That's probably what will swing it eventually: I didn't think about that. SDCs will keep improving and humans are stuck at the same level. Eventually the numbers will add up right and things will sort themselves out. Shame that so may

Re:

[gweihir]

You seem to be pretty uninformed about the current state-of-the-art and the current progress. That some companies are doing reckless things does not invalidate the technology. And, as I said, it is not there yet, but it is getting closer. There are no _fundamental_ limits, except for situations in which many humans struggle as well. Cognition is definitely not needed for driving. Also, there is nothing half-assed in the research being done on autonomous driving in the last 40 years. There is something half-

Re:

[Rick Schumann]

*sigh* keep believing in Unicorns and Fairy Dust all you want. 'Deep learning alogorithms' are NOT ENOUGH. There is precisely ZERO ability for this software to 'THINK', it can't, it doesn't, and THAT is why it won't work. We don't even know how it is any brain, human or otherwise """thinks""", it's still a mystery to science, and you don't seem to understand that -- and media and marketing hype have you convinced otherwise. That's all you're going on: media hype and marketing department hype. Meanwhile you

Re:

[gweihir]

Seriously, what is wrong with you? Your arguments cannot hold water. You are grandstanding on nothing. Do you have some trauma?

Re:

[Rick Schumann]

No, there's nothing wrong with me, it's you that clearly and objectively doesn't understand. Stop believing the hype and get some actual facts.

Re:

[gweihir]

I am an engineer and scientist. I do risk management as part of my job. I have first talked to somebody doing research in this area bout 30 years ago. I do not fall for hype. You are badly off.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[JaredOfEuropa]

What I’m afraid of is that deaths by autonomous vehicle will not be accepted by the public like regular traffic accidents. If a driver causes an accident, the perception is that it’s bad luck, unfortunate circumstances in poor driving conditions, perhaps a little overconfidence, etc. Happens to the best of us. But we won’t accept this from a self driving car precisely because we expect these things to do better than us meat sacks. Even if their fatality rate is only 5% of that of human dri

Re:

[imidan]

let's just say self driving semis the first twenty years in kill 800 people every single day in the US alone. That's still a massive positive versus the thousands that die with the non-self driving versions.

I'm not arguing against the eventual benefits of self-driving vehicles, but your numbers are way off. According to the NHTSA (https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812451), for 2014-2016, the number of people killed in accidents involving 'large trucks' averages 11-12 per day. If yo

Re: So tell me again

[slack_justyb]

You're right, it should have been 8 sans the extra zeroes I added from a slip on the numpad. Honestly, I should of proofread before hitting submit and I would have caught that. All I can say is that it's my own fault for having not corrected that before hand.

Re:

[gweihir]

And then the insurers will realize that despite the occasional "horrific accident", the driverless ones have far less of these per distance traveled. That will drive cost down even further.

Re:

[Rick Schumann]

More bullshit. Some so-called 'autonomous system' kills someone because it's half-assed and fucked up? More tragic, a human driver might have had a chance to do something illogical or creative that saved lives. Keep living in your AI fantasy world though.

Re:

[gweihir]

I am not in that fantasy world. Autonomous driving is automation, not A(G)I, no "creativity" needed. Standard "low energy system" and redundancy approaches from engineering are a lot better than what humans can do, when they eventually work reliably. We are not quite there, but it is getting closer. And the human driver is much more likely to be intoxicated or distracted or tired and to hence have created the problem in the first place. Most people are _bad_ drivers.

I have no idea what your issue with auton

Re:

[goose-incarnated]

And then the insurers will realize that despite the occasional "horrific accident", the driverless ones have far less of these per distance traveled. That will drive cost down even further.

Humans (drunk, sober, tired, texting, etc) average +-250000km before losing control of a car or having an accident.

SDCs in 2014 averaged +- 8000km before needing human intervention. Five years later, the number sfor SDCs haven't changed by much.

At the current rate of SDC progress, we're still looking at millions of years before they manage to average 250000km without losing control.

Since 2012 everyone and the AI dog has been telling us how AI for SDC is only five years away. The five years have passed, and

Re:

[SlaveToTheGrind]

I'm all ears.

Oh, Jeremy -- anyone who has read your endless stream of bitter, cynical, goddamn-the-man posts for any length of time knows that simply isn't true.

But back to the subject: These things are running on a single, short, controlled route at sub-walking speed, as many others have pointed out. Your logic is akin to seeing your child take its first bumbling steps and smugly predicting the demise of the current record for the 100-meter dash. Maybe you should leave that sort of untethered extrapolation to XKCD w

Re:

[gweihir]

They are already there. This is a niche, admittedly, because it is short-haul and speed is non-critical. But they are replacing 10 drivers with 1.

"up to 5 km/hr"

[Anonymous Coward]

So... walking speed?

Re:"up to 5 km/hr"

[gweihir]

That is not the important number here. The important numbers are "60% cheaper" and "1 driver supervisor for 10 trucks". Sure, this is a specialized application, but it will help gather data and make optimizations. And it _is_ mostly autonomous driving on a public road and the cars decide themselves whether they need help. Also, these things will get faster with more maturity. Every dike failure starts small and then grows. Does not make the flood at the end any less impressive.

Re:

[gweihir]

Fortunately, engineering does not care about your uninformed doubts.

Great news

[Mes]

Plenty of jobs being created by automation. Nothing to worry about.

Re:

[gweihir]

10 truck-drivers -> 1 remote operator. Pretty much the same rate Amazon warehouses do with robots. There it is 10 robots, 1 robot supervisor to solve the occasional issue. Of course, there is no stopping this, unless you want to go the route some religious cults have chosen.

US Manufacturer Levels?

[0100010001010011]

Has Tesla, Waymo and friends officially stated what level they are?

Rails

[Cmdln Daco]

Think how much expensive tire wear can be eliminated if they just install steel wheels and some rails.

Re:

[gweihir]

That is obviously not the case, or they would have them. Remember, this is Europe, not the US. Also, who says these could not be autonomously loaded on flatbed cars for long-haul?

Another Trade Harpooned

[Jim Sadler]

Considering the millions of people involved in truck and cab driving we best have plans on just how to support them as automation eliminates their trade. We simply can not maintain a come from behind approach and try to correct the problem until it overwhelms us. We need to get out ahead and create the systems that will allow the displaced workers to thrive in this world. The same applies to regions as well as individuals. For example W. Va. suffers greatly over the end of the coal era. Yet we have n

Re:

[iggymanz]

no, you're overreacting.

a 5 km/hour truck, all of 3 mph for you USAers... this is like an oversided delivery golf cart

real truckers meanwhile drive tens or a hundred or more kph

Re:

[110010001000]

West Virginia was never prosperous. What are you talking about?

Re:

[Joe_Dragon]

at least over the displaced workers can just destroy the thing that took there jobs and end up hear

https://www.mic.com/articles/1... [mic.com]

Re:Another Trade Harpooned

[hdyoung]

Yeah, when has that happened? Every singe industrial, economic, scientific or technological revolution that has ever occurred, the changes happened without any real planning. The resulting human wreckage was dealt with after the fact. Usually, it took like a full generation before people actually adapted. There is zero chance that humanity is going to actually plan for the coming AI revolution, or the driverless vehicle revolution, or the robotic revolution, or whatever else.

A conundrum

[AlanObject]

Driver-less vehicles?

How can we have a rational discussion of this without a reference to Musk to bring out the best in all the skeptics? But here we are.

Re:

[110010001000]

We all know that Musk promised robotaxis by next year. Guaranteed. Done deal.

Re:

[jezwel]

I suppose you could use a self driving truck for some routes not involving any backing into docks, unloading, or loading, or fueling. But honestly don't you still have to hire a human to be a back up and do other manual labor like coupling, and uncoupling trailers and other tasks? Why would a trucking company invest in this technology that cannot completely replace a human? Where is the cost savings?

Someone will be at the loading dock anyway and can do these jobs instead of the driver.

Re:

[jezwel]

So in the near future, for safety reason, I'd expect somebody to be manning a remote terminal that would watch over or supervisor the autonomous truck, similar to the way a military officer might watch over the flight of an otherwise fully autonomous drone.

Bloody hell, it's right in the summary

An operator, sitting miles away, can supervise and control up to 10 vehicles at once

The problem with driverless and safety

[Dunbal]

You want to mess them up completely? Go stand in front of it for a couple hours.

Re:

[ledow]

And find yourself charged with obstructing a public highway, with CCTV evidence of you doing so.

You can do the same to an ordinary driver, same outcome, nothing different about automated vehicles in that regard.

Re:

[Dunbal]

Oh you do it on the B roads not the highway. I'm sorry I just can't seem to figure out how to change this bicycle tire...

My point is I believe that the same features that allow self driving cars to not run over pedestrians can (and probably will) be abused to do naughty things to self driving cars... theft, hijacking, disruption, etc. Safety becomes a vulnerability. Whereas with a human driver you never know if the driver is going to be such a bastard that he might just drive right over you.

An operatior, miles away, ...

[aberglas]

The key statement.

The truck does not need to be 100% autonomous. It can be controlled remotely. From a comfortable office, possibly in India.

Tesla way ahead

[110010001000]

Tesla will have full self driving by next year according to Musk. Coast-to-coast autonomous driving.

Re:

[gweihir]

Hahahahaha, funny. And in 3 years, they can probably drive you to mars?
Musk needs to smoke less pot and listen to the experts I assume his company has.

Re:Tesla way ahead

[ledow]

I would rather trust a company whose vehicles are limited to 5km/h and are gently pushing for more only after extensive testing and supervision than a company that rolled out "self-driving" capability to a populous as a faddy feature that only really works at motorway speeds, avoiding all kinds of regulation by saying "Oh, that's the driver's problem".

I would happily endorse a driverless shopping car, golf trolley, kiddies ride, theme park / airport disabled person's carrier, etc. at low speed, long before I want to even be near a Tesla on the road even with a human behind it.

Re:

[gweihir]

Please provide a reference for "public roads" here. Because I fail to find one.

Annnnnnnnd.....

[flacco]

...it's been raped.

exactly what America needs to do

[WindBourne]

Right now, we need to allow AP trucks to go from warehouse to warehouse as long as they are within say 1 mile of the federal highways. Once they get too far off, for now, AP can not be trusted.