The Forklift Truck Drivers Who Never Leave Their Desks (bbc.com) 92
Forklift operators are using remote-control technology that allows them to work off-site, controlling their machines from afar. The BBC reports: [A]s Covid-19 spreads easily, the warehouses dotted along the world's supply chains have become potential hubs of disease transmission, says Elliot Katz, co-founder of Phantom Auto. Phantom Auto's technology is now installed in around a dozen warehouses in the US and Europe, he adds. Some of the warehouses using Phantom Auto's technology fence-off the space where the remote-controlled forklifts work, says Mr Katz and the forklifts are also fitted with microphones so the operator can be warned should something be about to go wrong. "If someone is behind that forklift and says, 'Hey, you're about to hit me,' the operator can hear it just like he's sitting on the forklift," says Mr Katz.
Among the other firms working in the teleoperation space is US start-up Teleo. It specializes in retrofitting construction equipment so it can be driven remotely. It has just started a trial at a quarry for an unnamed client. In this case, Teleo has adapted a large-wheeled loading vehicle so it can be controlled from an office on site. In the future, a driver could sit in the office and remotely control a variety of vehicles nearby. That might mean fewer people would be employed on-site overall but Teleo argues it makes the role safer for the driver.
But the idea of vehicles driven like this is controversial for some. There's always the possibility a terrorist, for example, might try to hack such a system and use a teleoperated car or truck to kill people. Mr Katz and Mr Shet [Teleo co-founder and chief executive] both say their firms have thought about this scenario and add that their engineers have introduced various steps to make a cyber-attack harder. For example, by encrypting communications between teleoperator and vehicle, requiring authorization of drivers and automatically shutting down vehicles should they lose access to a reliable communications signal. No-one can guarantee that such a system will never be hacked, though.
Among the other firms working in the teleoperation space is US start-up Teleo. It specializes in retrofitting construction equipment so it can be driven remotely. It has just started a trial at a quarry for an unnamed client. In this case, Teleo has adapted a large-wheeled loading vehicle so it can be controlled from an office on site. In the future, a driver could sit in the office and remotely control a variety of vehicles nearby. That might mean fewer people would be employed on-site overall but Teleo argues it makes the role safer for the driver.
But the idea of vehicles driven like this is controversial for some. There's always the possibility a terrorist, for example, might try to hack such a system and use a teleoperated car or truck to kill people. Mr Katz and Mr Shet [Teleo co-founder and chief executive] both say their firms have thought about this scenario and add that their engineers have introduced various steps to make a cyber-attack harder. For example, by encrypting communications between teleoperator and vehicle, requiring authorization of drivers and automatically shutting down vehicles should they lose access to a reliable communications signal. No-one can guarantee that such a system will never be hacked, though.
Pretty smart (Score:4, Interesting)
Release a dumb remote-control product now that addresses the challenges of COVID-19, appeals to management and keeps the unions quiet.
Then later, when the installed base is large enough and the technology is mature, replace the drivers with a simple software upgrade.
Nice trojan, nice thought-out business plan.
Re: (Score:3)
In this case you could as well have a completely robotized warehouse where no people are allowed entry unless special circumstances calls for it. At which time everything shall be shut down.
But when you have an operator on a forklift and not remotely then that driver can often figure out when there are problems with either the forklift or correct some of the load on a pallet before it's put up.
A problem with the forklift can be as simple as low pressure in a tire.
So it might be that you fix one thing and ge
Re: (Score:2)
Re: (Score:2)
The question will come down to how much will that Problem Fixer Cost, compared to just a normal operator. The pricing may work out, or it may not.
As a software architect I get paid 3x as much as people doing data entry. There are jobs that I can code automation for many of the data entry jobs. However to do the job correctly the coding may take 4 months. Only for the requirements to change completely the next year. So my effort in automation will not be a cost savings.
However if it is a case it only takes
The software does not need to be perfect (Score:2)
Version one is driven completely manually. May actually be safer than a driver on the machine as it would have a rear facing camera that does not require the driver to turn their head.
Version two is computer assisted. To move quickly down narrow aisles. This already happens in many warehouses. The computer also is a safety check, preventing the machine from running into things.
Version three the computer can do routine stuff autonomously. But a human still needs to be there for more complex tasks. This
Re: (Score:3)
It's not just a problem with the forklit. The pallet has a label that's coming off, or you aren't shipping a full pallet, or you need to shrink wrap it before sending it. Also maybe the cameras can be repointed, The other thing us unless things are put right on the racks and nothing shifts you might have to lean out or get off the lift and walk around to make sure it's not leaning weird or whatever. Do these forklift drivers just drive the forklift and they still have staff to do all the other stuff on site
Re: (Score:1)
Obviously not, they are offsite. But they can still do the "drive the forklift part" and have someone else do the rest. I'm in agreement with GP though, this is baby steps to mostly autonomous forklifts (already on the market). There will still be a remote operator in automatic mode to take over in the low percentage cases it calls for help, and still people onsite to do the afore mentioned tasks, at least until they automate those as well. "Lights out" warehousing is a when, not an if at this point. I
Re: (Score:2)
Fair enough. I suspect it's at the moment a less economical way of doing things, basically it's R & D spend to work towards the scenario where you don't have the $15/hr lift driver and still a $10/hr factory worker following it around for labels, wrapping binning etc. + all the extra cameras, computers and office space. Getting better but a lot of warehouses have really crappy wifi too so they'd need to make that way better to handle multiple cameras per truck. 5G-6G maybe a better way especially since
Re: (Score:2)
"make sure it's not leaning weird": Ships have inclinometers (https://inclinometers.com.au/shop/inclinometers/mode-15s/, or see https://inclinometers.com.au/s... [inclinometers.com.au] for the dual axis version). Wouldn't that work, in fact be more accurate than a Mark 1 eyeball?
That of course doesn't address the shrink wrap etc. problem, although somewhere in the past couple days I thought I saw a semi-automatic shrink wrapper designed for use with a forklift. Unless I'm hallucinating it...
Re: (Score:2)
What are you going to do put an inclinometer on each pallet? I don't mean the forks of the forklift but the load on the pallet, or say you didn't get your forks in all the way and you are only picking up the front of the pallet, or the pallet has broken boards so even though you are lifting straight things aren't lifting right, or the pallet is catching the edge of another pallet or the rack etc lots of things. Again the cameras help but sometimes you have to get off the truck and look at things from a diff
Re: (Score:2)
Forklifts have solid rubber tyres.
Re: (Score:2)
Not all of them, only some.
Re: (Score:3)
Then later, when the installed base is large enough and the technology is mature, replace the drivers with a simple software upgrade.
You skipped the step where the current drivers get replaced by operators in another country first.
Re:Pretty smart (Score:4, Insightful)
Unlikely. Moving heavy objects with maybe a centimetre grade tolerances if you're lucky requires near instant reaction time.
Latency is real, and as games have shown, cannot be bypassed by "the internet".
Re: (Score:2)
Re: (Score:2)
Problem being that this is where automation actually routinely fails. Automation is great for specific repeatable tasks. Getting the pallet just right in the median warehouse is the exact opposite of that - it's minute adjustments based on many variables.
To be fair, you probably could mount enough cameras with enough resolution and a powerful enough computer onboard and powerful enough software to manage these tasks adequately. At which point, you're going to have a golden forklift that no one will want to
cm accuracy has alreay been automated (Score:2)
Forklifts already drive themselves quickly down narrow aisles, often guided by a wire on the ground -- old technology. The operator just sits there, and the machine is much less likely to hit the narrow sides than a human.
Re: (Score:2)
Because that is the easily automatable part of the warehouse actions. Routing.
The hard part is picking stuff and placing stuff. That is where automation fails to this date, even in companies that invested billions in it, like Amazon. It's why Amazon has those special shelves and robots instead of normal pallets and forklifts where they can. Latter are not automatable in a cost-effective way with current technology.
Re: (Score:2)
The way to handle that is by having automation of just the low-level stuff. So the operator is still making decisions just above that level, like "now turn to point at where I clicked. Now move forward until your palette sensor makes contact," but not joysticking the lowest level decisions like "stop going forward....now!"
Workable, but, compounds the latency issue by adding additional moves, and loses the instinctive abilities experienced fork lift drivers bring to the task, When you have done any complicated job long enough, you do not rationalize steps - you look, you know and you do. Even amazingly complex operations become instinctive with enough repetition. The ability to immediately deal with the unexpected, seldom seen, the new, and the hidden exception (looks okay but something does not feel right - experience tell
Re: (Score:2)
Maybe another company that vastly under-estimates the difficulty of building a self-driving forklift robot.
They exist already but only for highly controlled areas with expensive positioning systems to help the robots and avoid needing to do any complex vision stuff. Which is why many places still use human drivers, they are far more versatile.
This will change the nature of the job though. One person need not work at a single location any more. If they only need a part time driver they could rent an hour or
Re: (Score:2)
Who cares how large the industry is? The only questions are what is the state of the art in robot forklifts, and are they ready to replace human drivers.
Re: (Score:2)
If you suck AmiMoJo any harder, he's going to explode.
I respect AmiMoJo dozens of times more than any AC shit-talker. Grow a pair, then come back to tell us all about what we should think.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I wouldn't call it a simple software upgrade. Currently we are barely able to keep automobiles to drive safely on highways. Driving a forklift is much more complex than that. It isn't a case of stay on the road and don't hit anything. It is a case of traversing an ever changing set of environments in a warehouse, on a parking lot, go off onto the lawn. Find the correct object, and target the forks in a precise location as to not puncture or damage the product and/or what is holding it. Bring it and mo
Re: (Score:2)
Driving a forklift is not more complicated than driving a car, which is by far the most complicated thing most humans do on a daily basis, and it also happens a lot slower. If the lift has to stop and think for a couple of seconds before it can pick something up, no big deal.
Threading the forks under a pallet is not even close to being as complicated as threading a car through traffic. For one thing, the pallet isn't moving while you try to line up on it.
I drive a fork lift around a gravel yard on the regul
Re: (Score:2)
On the other hand, in real life not all pallets are properly secured and it helps if the driver can notice that the load is shifting and gently stop to re-secure it rather than wait until it falls off. If the load does fall off, current robots are unlikely to be able to clean up the mess short of writing it off and bulldozing it into a bin.
Re: (Score:2)
Making a self-driving car is not that hard. Making it go fast is a bit harder, but the real hard thing is making it go fast in situations where it may encounter human traffic that also goes fast. All of this does not (need to ) apply to autonomous forklifts. Autonomous forklifts are a no-brainer, especially in situations where humans can simply be denied access, like most larger warehouses.
Also note that autonomous forklifts already are a thing. They're inevitable.
Re: Pretty smart (Score:2)
Re: (Score:2)
No need to connect automated warehousing to some conspiracy theory, it already exist. Do a youtube search and you'll find several vendors, here's a five year old example of a Belgian warehouse moving regular pallets with robots: https://www.youtube.com/watch?... [youtube.com]
Re: (Score:2)
s/simple/complex and unreliable/
ok, you did say when the technology is mature...
Actually, I don't know what the relative difficulty of driving on roads vs. fork lifting in a warehouse is. On the one hand, there's not a lot of traffic in warehouses, and the speeds are much lower. On the other, on roads you don't have to deal with three dimensions (at least until we get our flying cars, now very overdue) and a center of gravity that quickly and dramatically changes.
This seems like a fine idea, (Score:3)
Re: (Score:2)
Yup, hope their traffic is encrypted and there's some form of authentication, otherwise you're going to get a new breed of hi-tech theft and there won't be anybody around to prevent it. Fun times when Mallory can hijack Alice's remote-controlled forklifts and drive off with the goodies.
Re: (Score:2)
How about the liabilities for death. They own the remote control tool that killed, their liability for negligent homicide is theirs. If a terrorist hacked it and killed a bunch of people, the owners are still fully liable, they failed to secure the mass murder tool.
The law should be tightened up, your robot or remote control device kills, then you are fully legally liable, choose a perish, you murdered them or you failed to properly secure the device proven in that it was used to kill someone. That it got
Re: (Score:2)
If someone physically broke into your car and drove off and hit someone, the owner of the car is not liable like that, not even if the owner left the windows rolled down and the keys in the ignition. Why should the owner of a remote vehicle be more liable? I'm not saying you're wrong, merely that I don't see it as a slam dunk that you do, and I would like to hear more about your reasoning.
Re: (Score:2)
If someone physically broke into your car and drove off and hit someone, the owner of the car is not liable like that, not even if the owner left the windows rolled down and the keys in the ignition. Why should the owner of a remote vehicle be more liable? I'm not saying you're wrong, merely that I don't see it as a slam dunk that you do, and I would like to hear more about your reasoning.
Very apt comparison. I hope any remote operated vehicles are able to say who was controlling it to prove a hacker gained control. Laws also need to determine culpability of manufacturer in the case of a hacker breaking in.
Re: This seems like a fine idea, (Score:2)
RTFS
Re: (Score:2)
Re: (Score:2)
and you can just kill the power / LP on the unit or hit the E-Stop button.
Speed of light delay? (Score:2)
I'm curious how they've overcome the speed of light delay. If the operators are on-site, not a big deal. But if they're at their home miles away, there will be that delay between when they command the forks to life and when it occurs.
Then there is the question of force feedback. When you're on site you can feel how the machine works. You can tell if you're lifting too much, if the load isn't centered properly, and so on. Do they have sensors for this?
While they've thought about someone trying to hijack t
Re: (Score:2)
I'm curious how they've overcome the speed of light delay
Drone operators bombing the shit out of innocent civilians in Afghanistan from the comfort of their cubicle in this US seems to have worked that out.
Re: (Score:2)
I'm curious how they've overcome the speed of light delay
Drone operators bombing the shit out of innocent civilians in Afghanistan from the comfort of their cubicle in this US seems to have worked that out.
The payload of a drone doesn't need to be as precise in location as a forklift.
Re: (Score:2)
I'm curious how they've overcome the speed of light delay.
A forklift in Ohio operated by a driver in India would have a speed-of-light delay of under a tenth of a second. Compared to the speed of normal human reflexes, that is negligible.
Re:Speed of light delay? (Score:5, Insightful)
Actual round-trip network latency between Bangalore and Cincinnati is in the neighborhood of 225 milliseconds. Median human reaction times are on the order of 200-250 milliseconds. Doubling the reaction time is definitely not negligible.
Re: (Score:2)
Re: (Score:1)
this seems like one of those things where someone being on site is better than being remote
How much value would you assign to the advantage of being on site? Is it more than the value of not having a commute? Is it more than the value of flexible schedules and the ability to quickly replace a sick coworker? Or, as others have already noted, is the advantage of having you on site more valuable than the amount the company can save by paying someone who lives in a cheaper state or country?
Re: (Score:2)
Homes might not be too far away. Maybe there's on-campus housing. Or maybe they're living out of their cars and parked right at work, as has happened to multiple people in the high $$$ cities who cannot afford mortgage or rent. I mean, you're probably right, in most cases there's a problem, but it could be solved by having the remote operators live close by.
Re: (Score:2)
If you're operating a forklift and you get into a situation where 50ms makes the difference, you're doing it wrong.
Re: (Score:2)
The delay due to the speed of light is minuscule; for an operator 1000 miles away, it would be just over 1/100 of a second. Much more serious would be the computers in between, assuming the connection is over the internet rather than a switched network. I've been on Zoom calls where my microphone wasn't working right, so I call in with my cell phone. There's a quarter of a second or more difference in the two, with the sound on my computer's speakers (I think) arriving later. Very disconcerting.
Force fe
And so (Score:3)
automatically shutting down vehicles should they lose access to a reliable communications signal
Slowly, lest it be carrying a load, especially high up.
Re: (Score:1)
Your sig "(-1: Post disagrees with my already-settled worldview) is not a valid mod option." is wrong.
Only two people are allowed to use that mod option, CowboyNeal and this guy [slashdot.org].
Oh wait, that's you.
Sorry, should've checked first, you obviously already know about that exception. Carry on.
Klaus (Score:3)
people who have never been a forklift driver (Score:4, Insightful)
This is a solution from people who imagine what a forklift driver does.
Do you think they never need to get off their forklift? Pallets are never broken? Loads are never collapsing? Packages never break? Change the LP tank? That you never need to get off, walk into the truck, and look at exactly what you need to do to get that pallet out from under/atop that other one?
And that initial assertion "warehouses are becoming hubs for disease transmission" what? Last time I checked, commercial warehouses rarely have people within 40' of each other, much less disease spreading distances.
This is stupid.
Re: (Score:2)
You will occasionally need a human to intervene, so what? That still leaves robots or teleoperators doing the bulk of the work. So maybe it only takes 95% of the jobs... that will surely help 1/20 of the current drivers.
Re: (Score:2)
Yeah exactly. Also I know probably just a photo op but the two drivers in the pic in the article are probably closer together than warehouse folks are most days besides lunch breaks. Unless you let them work from home you can almost guarantee people working in an office will be closer together than someone in the warehouse/shop floor.
Another thing: irregular skids, ie having to move the forks around, or swap it out for a barrel lifter. I mean you can have a grunt around I suppose to handle that for the lift
Re: (Score:2)
"the two drivers in the pic in the article are probably closer together than warehouse folks": maybe, but it seems pretty easy to put a plexiglass partition between those workstations, and/or provide sufficient ventilation; plexiglass partitions between forklifts driving around, not so easy.
But I expect this will quickly become a wfh thing.
Re: (Score:3)
- Shipping 2 tons of freight on a pallet meant to take half that.
- Freight beyond the capacity of the forklift attempting to unload it (happens all the time)
- Oversize freight extending beyond the pallets
-
Re: (Score:2)
I believe you can tell pretty quickly who's actually been in the role, and who's just pretending.
Re: (Score:2)
Agreed. Reading most of the comments so far, it seems like most have never worked in a shipping environment before, understandably. But anyone who has worked LTL freight at some point will remember that people are idiots and the manner in which some material is shipped makes remote work impossible.
Not impossible, automated warehousing already exists but you're correct that packaging quality is a severely limiting factor.
Shipping containers (Score:2)
People used to say the same about shipping containers. It will never take off, freight is too varied.
Today, freight is made to fit into a container one way or the other with very few exceptions. And the docs are hugely automated. (And still very inefficient.)
Re: (Score:2)
I think you're completely correct it's just a matter of time (but I don't know that ports are inefficient). I do know that at some ports the union who organize the loading has made deals not to automate and some for retraning should they automate.
Re: (Score:2)
Nonsense.
While I can see the similarities between the situations, they're superficial.
IT MAKES SENSE to unitize international cargoes, where there are serious issues of security, as well as fundamental economics of stackability, loadability, etc involved and the scale is such that the cost-impact is ultimately small per unit to the consumers.
To do this on a lower scale - ie to 'unitize' pallets for automatic handling - never going to happen. The A/F industry has tried with their own containerization, but h
Re: (Score:2)
LP tank? I would have thought most forklifts would be battery operated by now. No?
Re: (Score:2)
It's about 60/40 electric today, the reverse of what it was when I started in the logistics industry and spent a lot of "quality time" at the wheel of a lift.
https://www.liftech.com/opport... [liftech.com]
Until fairly recently, electrics were nowhere near the performance of ICE FLTs; I don't drive them regularly now but I'd assume they're closer if not on-par today. I still wouldn't run an electric outdoors or in weather. LP trucks are much cheaper in the first place (although afaik electrics are cheaper over lifetime)
out source it to a place where the driver has noth (Score:2)
out source it to a place where the driver has nothing to lose or even put in an arcade game (don't do that or some kid may try to go for the death high score)
What about??? (Score:2)
More jobs (Score:2)
Now they need to make some with robot arms so I can actually work from home instead of coming in like a sucker.
You don't need terrorists (Score:3)
Have you ever worked around a forklift? If you have, you quickly realize that a forklift is kind of like a knight on a rhino dual-wielding lances. I've misjudged and sent a fork straight through a box, and the complete lack of resistance is frightening. Which is mostly to say I would not want to be anywhere around a forklift where the driver does not have full peripheral vision and I have no way of getting his attention and confirming he sees me.
if the warehouse is totally automated, and there are no humans on the floor, then this is not an issue. But as pointed out by another poster, things spill, loads topple, and machines break down. This means at some point people will be on the floor and when they are you need to consider how they can be kept safe if the operators aren't present.
Re: (Score:2)
A human on a fork lift can only look in one direction. A human controlling a fork lift through teleoperation can look at multiple camera views at once.
You could do that with the human on the fork lift of course, but why would you? It makes more sense for them to work remotely at that point.
Re: (Score:2)
But in practice, they can only pay attention to one of those views at a time.
Re: (Score:2)
They can only focus on one at a time, but they can scan many of them for unexpected obstacles. We do the same thing while driving, but also checking our mirrors. If they are set up correctly, you will be alerted of people passing or coming up behind you by the unexpected motion in your mirrors.
We have a 30' bus-to-RV-conversion and I couldn't even keep it in the lane reliably in some spots without splitting my attention between the road and the side mirrors. It's left mirror, road, right mirror, road, repea
Re: (Score:2)
Re: (Score:2)
Thinking out loud here: with teleoperation, the forklift is receiving commands via some network, presumably wifi (maybe wifi on steroids). Spill or loads fall off or machine breaks down. Human walks or drives over to the problem wearing/ carrying an override transmitter that causes any forklift within 50 feet to slow down and stop while the human goes by, then resume after the human is at a safe distance. Could that work? I suppose one issue is determining when you're within 50 feet, since the reception
Re: (Score:2)
Re: (Score:1)
Over 40 years ago I had a brother in law that was a fork lift operator for Coke. He doesn't have depth perception. Somehow he managed to do his job. Years later I learned how to drive one. I don't know how he did it.
Re: (Score:2)
forklifts are terrifying (Score:4, Insightful)
Forklifts are terrifying.
You might think I'm joking, but anyone who has worked in an industrial setting involved in moving pallets or heavy items know that forklifts are basically 2-4 ton boxes of steel that can turn in place faster than you can twist your body. They also have a pair of tines/forks that are basically like swords that could easily club you to death, skewer you, or cut you open if you catch a rough edge.
Skilled operators don't drive forklifts slowly. The most skilled operators of forklifts drive them fast - incredibly fast, and their forks move even faster.
The article is talking about remote forklift operation in special settings, but we are nowhere near the remote operation or automation of the majority of forklift operation, which occurs in close proximity to human labor, used in ways that are far too ad-hoc and way too fast to automate.
Re: (Score:3)
, but anyone who has worked in an industrial setting involved in moving pallets or heavy items know that forklifts are basically 2-4 ton boxes of steel
2-4 tons are the very small forklifts. Forklifts weigh about 4.5 tons on average and can easily weigh far more, as much as several American full size heavy duty pickup trucks loaded to capacity. Look at the weight capacity the forklift is rated for and then multiply that by 1.5-2x for the weight of the lift itself. Add to this the forks are several inches thick solid steel, it’s no wonder they go right through wood and sheet metal like a fork through cheese. They will turn flesh and bone to past
Re: (Score:1)
Time (Score:2)
Interesting. To jump ahead a bit, perhaps a nation may want to ban remote control physical robotic labor rom overseas, except perhaps for surgery or medical consultation.
OMG terrorists! (Score:2)
The dumb objections people come up with to new technology. It's not terribly difficult to hack a vehicle to the point where it can be driven remotely. The Mythbusters did it every other week. If a terrorist wanted a remotely operated forklift, they'd buy a forklift, a few other parts of Amazon, and make themselves one. It's far easier than hacking some actual industrial
Cyber attacks aren't thwarted by cyber security (Score:3)
...just like crime isn't thwarted by security guards. Security guards merely change the value proposition for the criminals.
Crime has always only every been thwarted by law enforcement. And law enforcement has only ever worked when criminals live amongst their targets.
Our problem today is exactly the same as the old problem of hauling gold bouillon on trains, across long distances, with outlaws around. You can't have enough security to make that much gold not worth the criminals time. And you can't catch the criminals later because they live in the woods.
The solution was to give those criminals a reason to live in-town -- because living in the woods sucks and living in-town is awesome. Hospitals, brothels, grocery stores, heat. Then you've got forever to catch them -- you can catch a criminal three years later. That's why the FBI became famous.
Today, we have cyber criminals "out-there on the internet", and high-value targets at accessible directly. No amount of security is meaningful because they can "get away with it". The keyword there being "away".
Figure out how to chase and prosecute them years later, and that's the value proposition changed -- they'll need to be running forever. Running from all the good things in life.
Here's a video of him at work (Score:2)
https://www.youtube.com/watch?... [youtube.com]
Forklift Wars (Score:2)