The automotive industry is rushing towards autonomous vehicles as a futuristic ideal. They haven’t got the autonomous part sorted just yet. However, as part of this push, the technology to drive vehicles remotely via video link has become mature.
In the United Kingdom, there has been great controversy on whether this should be allowed, particularly for vehicles piloted by individuals outside the country’s borders. That came to a head with a Law Commission repot published earlier this year, but since then, innovative companies have continued to work on remote driving regardless. Let’s dive in to the current state of play.
Why Go Remote?
Limited remote-driving technologies are already in common use, such as in industrial sites, mines, or the parking-summon features available on some new cars. However, full remote driving on public roadways is the key matter of interest of late, which would offer serious productivity improvements to a variety of industries.
A prime example is to allow the convenient delivery of rental cars. Typically, to deliver a rental car, a driver would have to drive the to the customer from the storage depot, and then make their way back by another means of transport. It’s a time consuming and inconvenient task. Ihe car could be remotely driven by a worker at a computer, this issue would be eliminated. Once one delivery was complete, the worker could log in to another vehicle and drive it to the next customer without ever leaving their seat. It’s not hard to imagine similar benefits for the trucking industry and other transit sectors.
New Rules Needed
However, the technology has drawn the ire of the Law Commission of England And Wales. The independent statutory body is charged with reviewing and making recommendations on the law of the land, and published advice to the government concerning remote driving earlier this year.
By the nature of the way vehicles used to work, there is currently no rule in the UK that drivers must be inside the vehicle they are controlling. This would naturally open the doors for remote control of vehicles on UK roadways. However, the commission has recommended that the government bring in new legislation to control and restrict remote driving. Key concerns surround the issue of responsibility. Presently, a driver who causes a harm out on the road can be held responsible for their actions. With the rise of remote driving, however, this becomes more difficult. Presently, if somebody remotely controlling a vehicle from outside the country were to recklessly cause a crash in the UK, it would be difficult or impossible to hold them responsible for the incident.
The report notes that there should be robust rules that ensure remote drivers are responsible for their actions just as if they were directly behind the wheel. However, at the same time, drivers shouldn’t be held accountable for issues beyond their control, such as a crash caused by a temporary loss of connectivity to the vehicle. International agreements would be required to enforce liability across borders. Failing that, the report suggests remote drivers from overseas should not be allowed to control vehicles operating in the UK.
The overall thrust of the report from the Law Commission is that while the technology promises exciting benefits, there is a need for new legislation to ensure it is used appropriately and safely. It also highlighted issues around cybersecurity, in which remotely-driven vehicles could be potentially hijacked by bad actors. Beyond that, it also explored the human factor—where drivers may inadvertently be less attentive or alert when driving a vehicle remotely. The review advised that new laws would be required to shore up these issues, while also making it definitively clear whether remote driving was legal in the UK, and on what grounds.
The State of Play
While the Law Commission’s report takes a very conservative approach, that hasn’t stopped innovators in the field from charging ahead with their work. Imperium Drive has been a leader in this area. The company has recently rolled out remote driving with its Fetch vehicle system after 18 months of trials in Milton Keynes. It’s not the only player in the industry, but it’s a forefront player in the UK’s nascent remote driving space.
Fetch uses cars that are remotely controlled by operators at the company’s head office. The cellular network is used to let the cars communicate back to base, with 360-degree cameras used to provide the remote driver a complete view of the road. The remote driver controls the car with a steering wheel and pedals, while taking in the view via an array of screens.
Customers can hail a Fetch car via a mobile app within four miles of the Milton Keynes city center. The car is then remotely driven to the customer by a Fetch operator. The customer can then drive the car themselves just like a regular rental car, with a remote operator then driving the car back to base when the rental is concluded. The company states that for short trips, the service is as convenient as a taxi or rideshare, while it comes in significantly cheaper for customers travelling longer distances. Customers will be liable for paying an insurance excess while they are driving the vehicle, just like any other rental car. Obviously, this won’t apply in the event the car has an issue while being remotely driven.
Imperium Drive hopes that in time, its system could replace the need for private car ownership. The company is working on full autonomy as well in pursuit of this goal. For now, it operates four electric cars in Milton Keynes, with plans to add city-to-city and airport transfer services in future.
The company has had to develop a range of technologies to support its remote driving system. While lag or dropouts are frustrating in an online game, when driving a real vehicle on the road, they spell disaster. Imperium Drive employs advanced algorithms to predict and assess the quality of its vehicle’s data links. It uses this information to prioritize sending the most crucial data for driving the vehicle safely. Beyond that, its vehicles have been programmed to gracefully handle signal losses and other link issues to avoid interference or network problems from causing crashes out on the road.
Going Forward (Remotely)
Industry groups and investors alike will welcome clear regulation on remote driving going forward. At least as far as the UK is concerned, it may require companies wishing to operate remote-drive services in the country to offer some kind of guarantee as to taking responsibility for any incidents caused out on the road. New legislation is yet to land, and in the meantime, the technology is seemingly outpacing the rules currently on the books. Expect that to change as governments around the world prepare for the rise of remote and autonomous driving in future.
> However, at the same time, drivers shouldn’t be held accountable for issues beyond their control, such as a crash caused by a temporary loss of connectivity to the vehicle.
“Beyond their control”, my hiney.
If you’re gonna put a vehicle on the road, then you’d better be damned sure you have control over it. If you can’t be appropriately confident that you won’t “temporarily lose connectivity to the vehicle”, you have no business trying to operate it. That’s no different from being sure that the vehicle is in safe mechanical condition.
Frankly, regardless of the breathless cheerleading that pervades this article, I don’t believe that vehicles *can* be operated remotely with comparable safety to operating them locally. Even if you could be sure of not losing communication with the vehicle (which you can’t), the driver loses too many sensory modalities… and also isn’t in a position to actually do anything after something goes wrong.
The right approach is probably a total ban regardless of which country the operator is in.
Some good points. I agree there should be no case where loss of control is an acceptable thing, and certainly it’s not something that should be exempted from responsibility.
The idea that you might mow someone down and then, by not being present, just basically watch them bleed out when a human driver could have done something to help, is something I hadn’t even considered. Grim.
Since it is impossible to guarantee 100% uptime through wireless networks, a reasonable solution would be to have some level of self-driving that can safe the car whenever it loses connection.
At that point, why even have the human operator?
If the car is capable of full autonomous operation, just go full autonomous.
Because failsafe is much easier than full autonomous driving.
You don’t need to navigate, you don’t need to maintain speed, you just employ local conservative collision avoidance to immediately slow to a safe stop on the side of the road until communication is restored. You might cause a jam but you won’t kill anyone.
Also, because the certification is easier. The failsafe mode is by definition already operating in an exceptional emergency situation, and so hopefully rare. Certifying the ‘default’ mode of operation is more rigorous because it needs to avoid causing that jam.
>You might cause a jam but you won’t kill anyone.
Such jams often result in pile-ups because the stopping wave starts propagating backwards faster than the people can react to it.
Because full autonomous sucks and isn’t workable in real-world conditions. But it could probably go along with traffic for a couple seconds while the connection re-establishes.
The purpose is eliminating humans who draw wages, so if the “robot” is actually a guy in a cubicle in India and he gets paid 1/100th the wage, that’s still a 99% success.
Absolutely; it’s like saying someone should be able to drive from the wrong seat or with the controls improperly adjusted, or a tint too heavy to properly see, or headlights burnt out… People with narcolepsy have to show they won’t lose control briefly while driving or they won’t be allowed to do so either, and that’s very close to the same as a temporary connection loss. And that’s something they can’t choose not to experience, whereas the remote operator definitely chose not to come in person.
“However, at the same time, drivers shouldn’t be held accountable for issues beyond their control, such as a crash caused by a temporary loss of connectivity to the vehicle.”
Just adding this quote a third time. Really an astonishing sentence.
Issues.
This is a euphemism widely used to avoid dealing with important problems.
We have identified some issues with the situation.
Let’s be more precise. If the system fails, someone is responsible. Possibly all the people who made it happen: driver, government, software and vehicle manufacturer. Also if the system can fail, it’s probably not a good way to control a lethal vehicle in public.
“However, at the same time, DRIVERS shouldn’t be held accountable for issues beyond their control, such as a crash caused by a temporary loss of connectivity to the vehicle.”
Thought I’d quote one more time just for luck, it sounds ridiculous, but maybe not, the CEO should. If there was a well written law making them personally accountable including potential for serious prison time then perhaps companies might approach with the kind of caution this problem requires. Sadly I think the chances of getting this level of culpability is unlikley so a ban is probably the best thing to do.
> “there is currently no rule in the UK that drivers must be inside the vehicle they are controlling.”
So Mr. Bean’s recliner would actually be legal to operate?
https://www.youtube.com/watch?v=LVSLLWXdKV0
Fecking brilliant 😂
> Presently, if somebody remotely controlling a vehicle from outside the country were to recklessly cause a crash in the UK, it would be difficult or impossible to hold them responsible for the incident.
But the vehicle would be insured, so the insurer would pay out (as they would be held responsible). There would also, presumably, be a log of who was driving and when (as this is required by UK law so that speeding tickets and the like can be issued correctly).
This article is so positive towards these remote vehicle companies it makes me hope they’re paying the author for PR work. All these companies have done is find a new, “innovative” way to pay taxi drivers less while making the streets less safe. The idea that you could safely drive a car over a cellular connection is on its face absurd.
It’s a classic example of a government wanting to seem high-tech and down with the cool kids in silicon valley and jumping on the nearest passing bandwagon regardless of which direction it’s going.
They’ll probably spunk a load of public money at a few tech startups (likely started up by their mates from the golf club) which will fizzle out in a few years and then on to the next bandwagon!
Yeah. “…which would offer serious productivity improvements to a variety of industries.” Translation: we can fire all the people who live near you and pay people in the third world a tiny fraction of the price, and safety be damned.
Will they consider that what they are doing is not only stupid, but in fact morally wrong and antisocial? No.
Hi, part time taxi driver here. I would love to be completely physically safe from passengers and to be able to park my car in a better position for work rather than at home. The improvements this would make to my mental health would be worth being paid a little less.
Hopefully, the safety issues can be resolved because it’s something I’d use.
“However, at the same time, drivers shouldn’t be held accountable for issues beyond their control, such as a crash caused by a temporary loss of connectivity to the vehicle.”
They absolutely should be accountable!! Dude WTF?
The _driver_ shouldn’t be accountable. The _company_ definitely should. Not really any different to automotive companies being responsible if the car has a design fault that causes an issue, rather than the driver.
Nope. They ideally both should.
As a driver, I’m responsible for the car I’m driving. If it’s a company fleet car and its tyres are worn, that’s my responsibility as driver. Etc.
Speaking of worn tyres; how would one check that before driving remotely?
TÜV is a thing and you cant tell me you check your car tires each time before you drive.
I think you’re mixing up “driver” with “passenger”.
Ignoring that this is a terrible and stupid idea for infinity reasons and that it’s already happening regardless-
What happens when a remote driver is drunk? Few pints at me home office, remote drive for some extra cash. No accountability because “connectivity issues” and what could possibly go wrong?
Force companies to keep accurate logging of all interface data, and then it’s easy to tell the difference between poor control skills and legit connectivity issues
This kind of obviously insane flailing around is only happening because some hugely over-valued companies (which are zero interest rate phenomena and only make financial sense when you can borrow infinite cash without having to pay it back) based their entire business model for future profitability on the emergence of reliable, safe self-driving cars.
It’s becoming clear that this is a century problem, not a five-year problem, and the interest rates are going back up yesterday. So they are freaking out. And pulling bone-headed stunts. I don’t know how the Uber executive office hasn’t been raided by SWAT teams already.
Considering you can go to a computer Fair and buy GPS, GSM and all frequency jammers for $20 that plug into the car lighter socket, presumably used by car thieves, what are the lightly hood of the car behind and in front doesn’t already have them fitted and have a spare set when they carjack you.
the alternative is to totally disable if no signal which may not be such a good idea. the are also looking at autonomous taxies without even the option to locally control them. they are just a motorised box on wheels and probably lock the doors “for your safety”.
pff. just wire the person controlling the car so that in case of an accident, he will receive a electric shock of a serious kind so he ends up in hospital, just to prevent him/her for getting funny ideas. Oh and having a local allowed drivers license. I would not want a person with a ‘murican drivers license piloting a car here in the Netherlands.
but yeah, what could possible go wrong???
I wish all self-driving, remote driving and hands-free driving vehicles were required to have a big red flashing warning light on the roof, at least until technology has matured.
The driver is working from India. Oops signal lost.
This.
Driver is working from Russia. Doesn’t need to lose signal, can just run over whichever political refugees they want.
Yep. No different than a ‘drone’…. Seek and destroy. Can you image? All those driver-less cars (remote or otherwise) are just guided missiles either on a mission or hacked to carry out a mission. …
Hmm… would they be allowed to drive in europe?
I mean, at least in Germany and Spain, the driving license of some countries is not valid, and for good reasons… Would you really want to be in a car that is being driven by an Indian… in remote… or would you really want to share the street with someone that doesn’t know who has priority in an intersection?
I know… I know, not everybody is in europe a class A driver… but stadistically…
I think stadistically it’s a bit like being able to post online comments. It’s mostly incompetent idiots in ridiculous cars/bumbling morons with stupid usernames barely in command of their vehicle, trying to drive their car/message to god knows where but the majority of participants make it home safely. Until the next journey… remote… idiot…
Branding austerity as futurism.
“Imperium Drive hopes that in time, its system could replace the need for private car ownership.”
Which is actually the whole point of this idiocy.
There may be some validity to this beyond the conspiracy crowd. Look at the FedNow effort in the United States. So many pro-digital currency articles out there not taking into account potential control it gives the government in what you can/cannot spend money on and to whom you may give/receive money from.
Slippery slope comes to mind. I’m considered an old person by most of the readers of this website and maintain a belief that driving on roads, intermingled with human operators, will never be amenable to full autonomy/remote control. I believe for the foreseeable future, such driving will be limited to trains and similar restricted/predictable travel.
The term “conspiracy theory” was coined by a man named Cass Sunstein, who was working public relations at the time to discredit people who believed the Warren commission wasn’t sufficient to explain the assassination of JFK. The ngrams for the phrase before that time contain only a couple hits, I believe in totally different context and one is separated by a comma.
The concept was to smear the idea that powerful groups have agency, thus foreplanning, as extreme to the point of ridicule. The idea that secrecy would be used in the case that others might have different interests and object to certain actions was right out.
Obviously certain groups do exist, have agency, and have the power to carry out their own desires. The straw man is that there is a singular “They,” a group that controls the whole world to an extreme level of detail, which is the mark of a schizophrenic.
Conflating this with the belief in any lower-case “they,” a group of people or interests with potency, of which millions exist in the world—the sleight of hand is pretty brilliant. It basically makes accountability impossible, and the implication that any powerful figure doing anything offensive does so with foreknowledge or planning is forever tinged with insanity.
In short, yes, powerful people want to micromanage people’s movements and wealth as they always have since the beginning of non-nomadic civilization. Tyrants are not yet in the past.
Is this gig-economy “innovation” that seems like a comedic nightmare from the film Brazil a part of it? I kind of doubt it, but it is a useful idiot responding to incentives.
Public transport ftw!
Doesn’t the UK require drivers to be insured? I can’t image anyone insuring such driving.
Isn’t full fly by wire illegal for commercial and private vehicles in the UK?
Hey I’ve seen this one before!
https://www.youtube.com/watch?v=19zeMqh-zqY
What could possibly go wrong?
I think that the self driving car should have to pass the same test licensed drivers do before they are allowed to drive.
I think I would want it to be a bit better than some 16-year-old with a learner’s permit. It needs to provably be able to drive more miles per accident than an average professional human driver, in all road conditions. That last part is the important part.
This is the kind of problem where you can get 95% of the way there and look very impressive, declare that the problem is imminently solved, five years max, and that computers are obviously inherently safer and more infallible. But if falling into that 5% might kill you, that’s still one in twenty.. Not great odds. And then getting the rest of that 5% is an order of magnitude harder than all the work that has already come before. But to match the safety of a human driver (humans are very good at what they do, don’t believe the anti-human propaganda) you need to get that last fraction-of-a-fraction of a percentage right. And that’s going to be next to impossible.
People just do not respect complex problems. The urge to oversimplify and redditize a problem is endemic to silicon valley. There will be no more moonshots until this stops, only embarrassments.
As usual, “AI” and “automation” often translates to “a call center in Bangladesh where people get paid cents per day to pretend to be computers.” It’s a proof of concept!
Liability aside (As the company should be 110% liable for allowing the product on the road) The real scary part here is user error. Eventually an employe is going to commandeer a remote vehicle that already has a customer at the wheel.
The hackability? Rent a car – get unlimited time hands-on with it so you can configure it for remote hi-jacking.
-1/10 – would never use this service. Uber is becoming a hard enough prospect with the poor attitude of drivers in my area of the world.