Forklift truck drivers who never leave their desks | Hacker News Reader

Forklift truck drivers who never leave their desks

bbc.com

140 pointstosh128 comments

> Besides, there aren't enough heavy vehicle drivers to meet construction industry demand at the moment, asserts Teleo co-founder and chief executive Vinay Shet.

Horseshit.

Carpenter here. I've been trying to get my fork certification since January. I've had 8 classes in a row cancelled. They're not expensive or time-consuming. We simply can't get training because there's none being offered.

The unemployment rate in the USA is dropping but still around 8%, the highest it's been since 2012. Many hundreds of my colleagues are out of work right now.

You don't need technology to solve this problem. Hire people, and train them. If only 1/2 of 1% of the unemployed people in America are capable and willing to drive a fork, that's still 60,000 more potential drivers.

As a former software techie myself, I think this company looks suspiciously like the classic 'startup' which is trying to insert their technology into a situation, rather than trying to meet a market need. When a company says they can't find workers, I don't believe them, and I especially don't believe them in 2020 for jobs that take literally 1 day of training. Nobody can be that incompetent at hiring. Post "free forklift training + a job" on your personal FB page and you'll have 5 people show up this afternoon.

Which they absolutely should be doing, of course. If Teleo succeeds, they're going to need a lot of skilled drivers. This is classic "commoditize your complement".

Teleo is YC-backed, and Phantom's founder was previously in YC with a different company. Cute submarine. They're not really lamenting the lack of workers. This is pure marketing.

Do you have insight into why training is hard to come by? Is it a recent issue due to the pandemic? If lack of training providers is causing a worker shortage, isn't it true what that person is saying?

When I got my certification many years ago, the training course was given on our site by the truck driver from the rental company, who brought us two trucks for a couple weeks of use. He unloaded them from his trailer, then came in and gave us the training course in a few hours.

So it could just be a weird effect of all jobs and businesses being affected in some way, because that company didn't seem to have a separate job for trainer.

There have been some major and very public industrial safety incidents in the past decade. Everyone has gotten much more strict about safety protocols, especially in the last 2-3 years.

There was a time, I'm told, when it was "Whoever knows what the levers do, you're on forklift today". Then it was "You need to have taken a class at some point". Today, reputable companies will check the expiration date on your certification card when you show up for work.

Nobody wants to risk a pile of OSHA violations if something goes wrong.

Aha, that's after my time. I did it 20+ years ago.

A big piece is certainly COVID-19. Training centers are mostly shut down for public safety.

I don't think that's the whole reason, though. Were companies truly that desperate, they'd find a way. Many training classes this year have moved to part-online, part-in-person, for example, which helps with COVID safety, convenience, and cost. But demand is way down.

Ultimately, I think it simply boils down to money. We've gone through the better part of a year where most big projects were shut down, and if they were lucky, restarted months later, slowly, with restrictions. When you put a damper on every part of the economy (except Netflix and Purell), everything suffers. As a wealthy friend of mine once said: it's all about cash flow.

Reduced demand, few open training centers, DOH class size limits, limited spare cash from unemployed workers for class/book/cert fees, trainers who live elsewhere and might not want to get on a plane, etc. It's like hitting the breaker in Jurassic Park. You don't just flip it back on and get all the lights at once.

(I'm not saying we shouldn't have hit the pause button. Public health is critically important. I'm just saying it should be no surprise that labor and training got hit badly by it.)

It's hard to imagine a situation where it would not be worthwhile to have a trainer spend 8 hours teaching and certifying one new driver, if any real demand existed. Guess how much a good trainer costs for a day -- a single pallet often holds material with greater dollar value than that.

I don't know a lot about fork-lifting but maybe hiring is difficult across all domains. If you hire 5 people, and 4 of them turn out to be not good enough, you just wasted time and money. So if you need 5, you'll need to post for 25 persons. To train them you'll need multiple people and even more money...

A basic fork cert is like 2 days. It’s so cheap a lot of companies require you attend training they pay for and don’t go by old certs at all.

Hiring labor isn't like hiring a CEO or a doctor or an architect. Not this kind of labor, anyway. We're talking about a one-day safety class, with no prerequisites, as mandated by the government.

Imagine you were hiring for a job where one of the qualifications was "holds a valid driver's license". You wouldn't bring in 5x as many candidates as open positions because you assumed that most would have failed at that.

why dont you go back into software?

Not the person you asked, but although I'm in the tech field now, I worked as a shipyard welder for a couple of years after the first .com crash.

It was, by far, the most meritocratic, friendly and diverse workplace environment I've worked in, compared to my computer-related jobs. Payed well, too. If it wasn't for the (literal) toxic environment that is unavoidable in ship-welding, I'd likely have stayed longer.

"No sweeteners will cloak some forms of bitterness. If it tastes bitter, spit it out. That's what our earliest ancestors did."

> Hire people, and train them.

Training seems to be a dirty word in companies these days.

It's a cost all companies want someone else to pick up, either the employee themselves or a previous employeer.

There are still some companies out there that will do it though, eg Fidelity's Leap program [0] where they'll hire you, train you for 3 months give you 3 months as a team project then place you in the company. I went through it straight out of college and I'm still with the company today. It's been around for 10+ years now too so it's not just a fleeting initiative which is pretty neat.

[0] https://jobs.fidelity.com/page/show/leap

When ever you see a company claim "There are not enough workers"

what they really mean "There is not enough workers willing to work for the substandard pay we are offering"

I immediately envisioned an Ender's Game scenario where someone, somewhere is playing Euro Truck Simulator 2030 and, unbeknownst to them, is actually driving a real truck and making real deliveries.

I know that wouldn't quite work since absent the military academy aspect they might do things in game we wouldn't want done in real life (i.e. driving headlong into traffic because you missed your last delivery window and so are starting over), but the idea I like.

Introduce a ranking system where rank increases with successive perfect runs. Connect the top 1-2% of hardcore truck sim players to real rigs?

I thought perhaps you could substitute AI driving when the human does something egregious, but realistically it should work the other way around. AI can handle the rudimentary stuff, it's the extracurricular road situations you will want the human in charge for.

I would be curious if these simulations can be made accurate enough to filter out candidates for driving. Set up a series of scenarios and grade how each driver performs. Then long term compare that to actually perform if hired.

Building off another comment, enough drivers could be used to help train an AI how to drive or how to reach specific situations.

I don't think that will work. Either you need to have the players legally liable for any damage and injury they cause, or you have created a platform for motivated psychopaths and terrorists to kill people with impunity.

The thought experiment (based on that ender reference) assumes nobody knows their actions are being mirrored in reality.

I know, I just don't think that's feasible. The outcome would be some people figuring it out, and having legal cover to commit murder since they were just playing a game. Cue long legal battle over culpability.

That's what the America's Army games are for. Those are a project of the U.S. Army's recruiting command.

America's Army the original one ended up being a terrible game for playability, because it turns out making rifle shooting even semi-realistic is hella boring lol.

Lot of slowly crawling around the map on your belly, calling positions to snipers and grenadiers, and just waiting.

I haven't looked at it in decade++ though, so maybe it's an actual game now.

That's the point. They want to filter out the recruits that don't have the skills and patience for good infantry tactics.

Ah good point

Also I vaguely recall they made you go through an AQT, and like sit in a classroom and learn medic triaging, etc.

Ok I looked up the more recent version of the game, apparently it has in fact capitulated to populism and no longer emphasizes realistic tactics. Supposedly many servers will kick you for camping now.

I don't think AA was ever legitimately meant to find people in that way and was just a more traditional advertisement/propaganda tool meant to feed anyone into the recruiting pipeline.

Perhaps many players given virtual control of the same truck could provide some sort of consensus that filters out the undesirable behaviour?

I am looking forward to Twitch drives heavy equipment

“The Gang drives heavy equipment”

Sure, until 4chan gets involved.

"pass on the left" + "stay behind a car" would turn to "slowly plow into the left half of a car"

You'd probably need some kind of median filtering or voting as well as averaging.

My initial reaction is that this will not become very popular.

Many, probably most, jobs that include operating a forklift are not 100% just driving and operating the forklift, you have to get on and off it a lot, for example to change the propane tank or battery, hit the start button on a skid wrapper, fill out some paperwork, etc...

Many tasks that forklifts do could be automated to not include a forklift at all with some upfront investment. For example I used to move hoppers with raw material from one area onto the machines when necessary. They could have easily installed silos to store it with vacuum tubes to move the material around. But ultimately you still have to load and unload semi trucks, so if you already have the forklifts, you already have employees trained to drive them why automate it (well it would have made our jobs a lot easier, but I don't think that was a priority for them).

Are forklift jobs that dangerous to the operator that they should be automated? My impression was that the biggest risk involving a forklift was to nearby pedestrians. I do think the constant 360 view that those monitors provide probably helps; you also lose something in not having depth perception.

I'm very skeptical of the claim that more people would take forklift driving jobs if only they were from an office chair instead of actually on the forklift.

I think the best feature is your potential pool of workers is no longer limited to the immediate geographic area. The best case I can think of is a plant where the forklift only needs operated infrequently, so you have a pool of drivers that takes remote control of a set of forklifts only when needed, so their time is better utilized. But then it goes back to what are the other plant employees doing while the truck is being unloaded. I thinmk the b

The biggest risk involving a forklift is probably to building, equipment, and material damage. A ROV forklift probably would not improve that.

Kind of puts a whole new spin on the amusing Staplerfahrer Klaus training film. Turn it into a video game where the students are rewarded for maximizing carnage, and they won't forget those lessons easily.

https://en.m.wikipedia.org/wiki/Death_Race_2000

https://en.m.wikipedia.org/wiki/Carmageddon

First thing I thought of when I read 'forklift drivers' was Klaus. Here's a version with English subtitles:

https://www.youtube.com/watch?v=TJYOkZz6Dck

LOL.

Just so anyone else clicking the link knows what they're getting into, the video has a lot of comedy gore.

I'm very skeptical of the claim that more people would take forklift driving jobs if only they were from an office chair instead of actually on the forklift.

I agree, and in fact would go so far as to say that the physicality of the job is what makes it less boring and thus appealing to some.

Exactly, I have a friend who is a forklift driver for Costco. He loves that the work is very self directed, that a portion of the warehouse is essentially his to make as clean and organized as he wishes. The main requirements are that all the incoming goods be stored and all the empty bays restocked.

"I like stacking crates"

https://www.youtube.com/watch?v=D04wb7P_v-4

Have you introduced him to the game Wilmot's Warehouse http://www.wilmotswarehouse.com ? Under the covers, I consider it to be a word/idea game.

What does that mean?

The game is about putting boxes with icons or (abstract) shapes on them together so that they can be found easier later.

Consider the situation that you've been putting "things that are cold" together. Snowflake. Snowman. You've also been putting "Buildings" together. House. Castle. ... And now you've got an igloo. Where do you put that? "Things that are sports like". You've got a football, a baseball bat... and now a hockey stick. Does that go in the cold things? Oh, here's a stadium - does that go in the sports area? or the buildings group? You've been putting round things together - frog eggs, wheels, and this disk with a slash through it. And now you get things that look like pills... was that thing with a slash through it a pill? Should you move it over to the pills section? or leave it with the round things? Or maybe move the pills adjacent to the round things.

So on one hand, the game is about organization of things... on the other, it's a word and idea game about how you group particular things together.

yeah who wouldn't want to drive a forklift even just for fun?

I think there's so much more to it. once they're remotely operated is quite easier to install an ai in the control loop.

it will first be training only. it will then start taking off the less risky task, like driving the lift to parking space after being used. Human will be gradually moved into facilitator jobs, taking over the difficult situation, and ultimately supervision only

however I am left wondering why, in the midst of the autonomously driven vehicles hype, people are focusing on the messy, unpredictable road environments instead of working on these kind of in-between situations that are harder than an automated warehouse but simpler than the public world.

This is such a silicon valley response. In what way is it easier to add actuators to a forklift, and get a second set of control hardware that connects through a network (And all the additional problems that brings in) than to just put sensors on the vehicle?

because the net would have to control the actuators anyway, but it's not a given that an actuator can move controls as fast as an human would be able to. putting the human in control of the actuators instead of sensors on controls solves that issue.

> to change the propane tank or battery, hit the start button on a skid wrapper, fill out some paperwork,

These are all much lower skilled jobs that are easier to staff.

There's a pretty low barrier to entry to driving a forklift, at least in the jurisdiction I did it in. Training took place on the job site, by other employees, and it took no more than a few hours.

I know a company that does teleoperated food delivery using drivers from Columbia, to save on salaries.

At $4/hr, there's no big rush for self-driving.

That could apply to forklifts too.

Guess you mean Colombia.

So this is what the Tesla self driving beta is all about.

Who needs AI and all that when you get somebody from a third world country at low pay and qualified by completing GTA V.

Haha I have been driving for 27 years but can't get past the third level of GTA V)

Where is this? I would expect that teleoperated vehicles on public roads are not legal in most jurisdictions.

The people I know who are forklift drivers are doing so because they don't have other skills and getting trained to do forklift was their only way to get a living wage.

I really don't think those people would pass on the job if it moved to an office chair.

Drove forklifts and reach trucks for 4+ years. Often times there is a bit of finesse needed to shift a pallet into position for you to be able to pick it up cleanly and safely. Not saying it couldn't be done from a desk control, but without trying it out I'd say something is lost not be onsite and seeing the full impact of what you're doing.

Agreed, there is a "feel" to it in many cases. Starting to move something and feeling the machine shift or strain (indicating something might be stuck), or feeling that a load is possibly off balance, etc.

It's kind of like racing games. You can be a really good driver, expertly trained in maneuvering, stunts, tons of track time. First time in a video game, you are still going to crash the car 100 times, because you have no feel, and no depth perception. The more of a simulator the game is, the harder it will be to master.

I've driven probably 50 laps around Sonoma Raceway with 100+ other cars on the track without hitting anybody. I am fucking lucky to make that happen in a simulator.

A desk operator would have as much or better visibility as an on-site. They'd develop a different intuition.

People who are good at fighting games aren't better in a real fight.

This reminds me of Quad/Graphics in Wisconsin... I toured their facility the 90s. If you read pretty much any magazine, Quad/Graphics probably printed it. They had a gigantic 10+ story warehouse full of everything they had ever worked on, and automatic fork lifts on wires to fetch or put away anything. We met the founder Harry Quadracci, and it seemed to be the thing he was most proud of. They had printing presses the size of a school, and the warehouse dwarfed them.

A fire destroyed everything in july, 2002... 1 person waiting in a car outside was killed when the building collapsed on them. 2 weeks later later, Harry was found dead in pine lake under the pier on his property. It was deemed an accidental drowning, but if you could see how much he cared about that factory, you'd probably question the coroner. I heard a rumor he had rocks in his pockets.

Thanks for sharing your story.

It sounds like the "automatic fork lifts on wires" might have been the automated shelving system, which were found at partially at fault for igniting the fire [1]. A failure mode I had not considered before with automated systems. While in this case it was in an installation defect of a bad weld in shelves and not the software-driven hardware itself, with automated ultrasonic inspection of welds post-welding, could it have been prevented?

[1] https://www.insurancejournal.com/news/midwest/2007/10/01/839...

Is Phantom Auto carrying $63M in liability insurance for every installation of their technology?

Sounds like a good idea, forklift driving is a dangerous job: https://m.youtube.com/watch?v=ChOHnSL7ZCg

Did you notice how the driver is the only one not getting hurt in that video?

About 6:55 in I think he gets a little scratch.

Try watching til the end...

And how do you think the the reduction in situational awareness that comes from remotely operating something from half way around the world will make that better?

It's a joke, watch the video.

I don't need to watch St. Klaus the Patron Saint of Bike Shedding skewer someone to know that internet commenters who have mostly never operated this kind of equipment are mostly not qualified to discuss the practicality of doing it remotely.

Edit: and I was right, it was the Klaus video.

I spent 2 years working for a company (https://www.jbtc.com/en/north-america/automated-systems) that had been selling fully automated forklifts since the 90s (You read that correctly, 90s). They built special custom forklifts that were purpose built to not have human operators. It would have been awesome if the company had a way to call on human backup to drive the vehicles remotely when something went wrong or a special order came in. The more people you can keep off the factory floor the safer it's going to be.

I am not sure how well this would work, as others have said when you drive a forklift you really need to keep your head on a swivel. Both for general operation and of course for safety.

I got to drive a forklift during my first job right after high school in a shipping depot - I had no training but eventually they just let me do it!

One of the forklifts had longer forks and I didn't realize I was driving it. I tried to raise the forks and it got stuck on an edge of the shelf, eventually making a loud BANG and bending the shelf! Luckily the other workers just gave me a hard time but I didn't get in trouble.. because the boss made the same mistake all the time :D

Jeez, paint the steering wheel red or something?

That would be a great idea!! Maybe paint the entire thing red...

Is this also a solution to edge-cases for autonomous vehicles? eg. driven by AI 99.9% of the time, and for the 0.1% (eg. breakdown, flooding, police action) they are taken over by a remote driver and driven slowly and carefully back to normal road conditions?

Maybe some AI and sensors would also be useful in the fully remotely driven vehicles to ensure that people can also operate in the warehouse without too much fear of being run over.

Otherwise I'd hate to rely on somebody who-knows-where noticing in time that I'm there or responding in time to my voice.

I have long imagined that this is how long-haul trucking will be end to end computerized, as the highway miles are the (relatively) easy part to automate with FSD, as most of that tech is already here (adaptive cruise, lanekeeping). Getting the thing from dock to highway, and from highway to dock, will require (remote) human eyes and hands for some while yet.

The future might look something like this https://www.google.com/search?q=spmt+remote&tbm=isch where highway miles are done by computer and the last few hundred feet are done by remote. For low speed work, sitting in the cab is a significant disadvantage.

I just bought an electric car and it doesn't come with a spare tyre - the battery takes up too much room. If you get a flat you just have to call roadside assistance. So there are already good networks set up to handle these edge cases without having to build in a remote driving system that hardly gets used.

I recently bought a Hybrid SUV. No spare tire because of the battery. But it comes with run flats, fix-a-flat, and a portable air compressor. Once the warranty expires I plan on mounting a real spare on the roof. We go camping and I just can't see roadside assistance covering the countryside.

True but with an AV they would presumably be geofenced to areas where they can be supported. I don't think it's likely that a remote operator would take control to drive an AV when it's outside a defined ODD (Operational Design Domain).

Haha. Run-flats aren't just a problem in the countryside. My wife's BMW doesn't have a spare. Run-flats are only good to ~50mph and 50 miles when flat.

So what happens when you blow out the sidewall on a Friday evening on an interstate in suburban NY? You call AAA, wait 3 hours in the pouring rain, and get towed the nearest tire dealer (that's already closed for the day), check into the hotel across the street, and then miss out on the Saturday activities you had planned while you wait for the tire shop to open and get your tire replaced.

I'll never buy a car without a spare again. Run-flats are nearly useless unless you just happen to be near a tire shop during open hours.

That's not particular to EVs. Many cars (could be most, I don't have anything aside from a lot of anecdata however) no longer come with an actual spare. Just a can of fix-a-flat and an electric pump.

I'm doubtful. The problem we see currently is humans are terrible at paying attention to boring things. And AI driving is largely boring. Right up until something unusual happens, but if the human is asleep/watching their phone/etc, it's already too late.

I don't see how a remote driver solves that problem. They're probably even more likely to be distracted or asleep because they aren't in the car and their life isn't on the line.

That just means that humans cannot be trusted to notice or respond quickly to an emergency situation. The situations parent listed do not require immediate intervention. When the system detects an anomalous situation it can raise an alarm and shutdown while it waits for a grumpy human to show up and resolve the situation.

and importantly if it were a remote human who was alerted and put in control they'd be waiting to do it as their job and not the sleeping/drunk/texting driver in the vehicle

I think there are a lot of other interesting applications here too:

* "drive-by-wire" optimization of operator inputs to save on fuel or reduce machine wear

* failsafe gps boundaries to prevent the chance of equipment leaving the operating zone

* smart personnel and equipment tracking to avoid collisions or accidents in the workplace

* future reduction in cost of remote-only equipment - cabins along with everything else needed for "creature comfort" requirements is probably a notable percentage in bill of materials and related shipping/running costs.

* cheaper labor operating costs - jobsite risks, repetitive-stress-injuries and operator insurance would go down for sure - it's also safe to assume savings by employing drivers from anywhere vs localized and often remote

* uptime: machines can run 24/7 with the exception of fueling - no downtime to change drivers or have a driver ferry back and forth for a shift change

* easy path to automation, even if extremely limited in scope during first iterations it is reasonable to see how simple tasks such as repositioning and repetitive tasks could be automated, even with macros.

* Having deeper telemetry about equipment operation (how many lb-ft has the forklift lifted in the last 3 months, how much time did the loader spend idling, how quickly did the machine respond to X input, what are the outliers for this machine vs all the other models in the fleet) can lead to much improved maintenance and preventive repairs. Maintenance programs in most cases today are simply based on hours and only the manufacturer or dealer would see anything beyond that with regards to machine usage.

Larger manufacturers seem to be exploring this and offering remote operating consoles as an add-on but it seems pretty far behind.

For a forklift operated in a perfectly paved area and loading/unloading at a fraction of its max specs, it may work just fine.

For forklifts operated on even minimally uneven grounds or for heavy loads (or for other kind of machinery such as a wheel loader) allow me to doubt it can be working without some kind of (sophisticated) feedback (or additional safeties or other automations).

Forklifts and wheel loaders provide feedback to the body, any professional driver will tell you how he/she "feels" the machine through the seat, my guess is that you can achieve the same result only through a somehow more "immersive" kind of setup than a couple monitors and a steering wheel.

Then, there is the issue with other machinery, or obstacles or people around, in person you have effectively (on demand) 360° view, having multiple cams (front/sides/back) on a same screen doesn't seem to me like providing the same as that (but possibly people can be trained for that).

I have also some doubts on the precision/speed of operation, i.e. I suspect that - missing an overall perception - manoeuvring will be either less precise or slower.

Remote control (but "local" one, with the operator which is "on site" a few tens of yards away) is already in use since several years for waste (landfill) compactors, but essentially the machines have no obstacles and more or less only run up and down.

A friend of mine is permanently disabled because of poor communication between excavator spotters. Had the ground not been soft that day he wouldn't have survived being run over. I definitely welcome innovation in this space.

What is the current upper bound for camera drone flight time? Is it feasible to have these operators park a few 3rd person POV drones to switch between? There is only so much you can do with views from the equipment before you need ground-truthing.

I can see that working with a rotating crew of light quadcopters. Having an "eye in the sky" would help monitor people and equipment on a site.

Or just have people drive the forklifts

I fear that forklift driving through a computer monitor will result in even less situational awareness than being in the forklift, thus increasing the risk of accidents.

It would be easy to argue that more accidents but with fewer human injuries is still a net win.

While many forklift-caused deaths involve the driver dying when, for example, a forklift rolls over; forklifts cause thousands of severe injuries each year to people around the forklift.

Those people would be threatened and injured even more by a driver that has limited spatial awareness because he/she is not there and is just watching a computer screen.

We would see more accidents but with fewer human injuries if automated forklifts would enable a human-free facility, but it's not going to be at that level yet.

I'm talking about excavators, not fork lifts. Earth movers. There are many jobsites where the only people on the ground are spotters. My friend was a spotter when he was runover by an excavator.

https://www.youtube.com/watch?v=7tdfizoornI

It’s a warehouse, why not take the simple approach of fixed cameras with intelligent video stream selection based on location?

Did anyone tell these drivers that their work is being recorded in order to train an AI? /s

Wondering why they are sitting in front of a PC with a helmet :-D

Probably OSHA guidelines not keeping up with technology. Technically they're operating machinery on-site. There's probably a line somewhere that states that if you're operating machinery on-site you need to be wearing the appropriate gear.

The concept of driving a vehicle while not being in the area wasn't a possibility they considered while writing the regulations.

And yes, you could say that they should "just use common sense", but that's a fuzzy line. There's often good sounding excuses as to why one should not follow the regulations to the letter "this time". And those are the times bad stuff happens.

Why the sarcasm tag?

That seems like exactly what Phantom Auto should be doing if they aren't already.

But is it be what the truck drivers should be doing?

I mean automating things may sound noble, until it affects your own job.

And wouldn't you be wary if someone came into your office, installed a camera, and told you your job would be fully automated in a few weeks?

[1] https://en.wikipedia.org/wiki/Forklift_Driver_Klaus_%E2%80%9...

WARNING: Possibly unsafe for work because of splattering and gory OSHA and other violations(of taste)!

[2] https://www.youtube.com/watch?v=jbInxdkHxk0

but 7.8 of 10! [3] https://www.imdb.com/title/tt0289477/

How long before these jobs outsourced to lower cost countries?

Pretty much never.

People who are basically commodity steering wheel holders will get outsourced to low cost countries and then to AI. People who's job requires thinking don't have much to worry about.

Most forklift drivers have duties that include a non-trivial amount of getting off the lift and doing some amount of moving things and a good amount of planning (often specific to the business situation that day) as to where stuff will go. Same goes for someone operating a front end loader or excavator. Anyone can hold a steering wheel. The reason they're not paid min-wage like delivery drivers is because they're paid to think and plan ahead. It's really hard to outsource that kind of stuff in a manner that is actually cheaper in the long term. I suspect that just like in IT, the companies that successfully outsource will be the big ones that heavily invest in profit to the point where any given person is just following process, not really thinking and planning (think mines, major manufacturing operations, large food distributors, some LTL freight, Walmart, etc).

> People who's job requires thinking don't have much to worry about.

Think again.. white-collar jobs will erode as AI becomes more capable / integrated into business workflows. Personally, I'm working on projects that basically capture the domain knowledge and experience of analysts and engineers, packaging it into ML models.

As soon as some rabid congress-person undermines OSHA.

I think a lot of countries and/or trade unions have that pretty regulated, so it would be interesting to see how someone sitting in a far-off country would get the necessary permissions.

And I use "interesting" in the Chinese proverb sense, as I would fear that a bit of lag or network outages would just Klaus[1] everything up.

[1]: https://www.youtube.com/watch?v=ChOHnSL7ZCg

Doesn't latency(network propagation delay) make that pretty much impossible?

70ms between Kansas City and Mexico City. I think that's sufficient for remote control of a forklift.

Incidentally, it's the same as network propagation delay between neurons in the brain and muscles in the hand.

But whattabout latency over satellite communications for drone pilots ?

Finally a meaningful comment.

I guess the many hours spent playing Street Cleaning Simulator 2011 [1] finally is paying off. I wonder if the Goat Simulator[2] will ever find a usage...

[1] https://www.tml-studios.de/game/kehrmaschinen-simulator/?lan...

[2] https://en.wikipedia.org/wiki/Goat_Simulator

Getting a really "training your replacement" vibe from this ... if you can do it from behind your desk, someone can do it from another desk ... perhaps I'm too cynical?

I worked with a guy during a brief stint in construction, and this fork op was so gosh darn good, i said to him; “you dont drive a forklift, you wear it”

This guy would be freaking having full on conversations while driving this thing around the factory, and spinning and be totally aware and conscious of all his surroundings - hes fn amazing.

He told me he has been driving forks since he was 14 years old - and hes like 45 or so now...

I was always impressed by his precision and skill.

For example, this guy: https://www.youtube.com/watch?v=-k3mVnRlQLU

The forklift is a 2m10s if you want to skip straight to it.

Teleoperated vehicles need a separate, completely airgapped pedestrian detection and avoidance system that can cut power and apply brakes.

One hurdle with forklifts is if you brake too hard then anything on your forks goes flying off the front.

I wonder what trucks will look like if they don't have to accommodate people.

Maybe this is a stepping stone to FSD - where we outsource drivers to low wage countries?

I'm imagining these tasks being put out on Mechanical Turk

Nice, bit Would want many monitors and hopefully VR soon.

But the drivers in the photo are right beside each other - with no covid distancing.

Looks like these guys are just sitting around playing video games...haha

The guy sitting in the background has a crazy setup.. 3 displays, maybe more above them?

Reply on news.ycombinator.com