Is there a reference design/concept for zero-human warehouse operations? Stretch looks like something optimized for warehouses designed around human operators. If humans weren't even part of the design consideration I would think the robotics would be less of a replacement for the humans walking the aisles and picking stock and more integrated into the storage infrastructure itself.
This is an excellent point, and probably makes this more of a transition robot (transitioning operations from human centric to robot-centric). Come to think of it, most Boston dynamic robots I've seen have been designed to operate around humans (and this probably makes them more 'friendly' to humans)
Probably the closest thing would be container harbors, where automated vehicles (AGV's) have been doing the grunt work of carrying containers around for nearly 30 years now (see e.g. https://vimeo.com/259128459). IIRC humans still operate the cranes though.
But yeah, assuming infinite space, I'd work on standardizing packaging for products; miniature containers, each containing one of a product. Probably would need multiple sizes of containers though. But a manufacturer would, instead of packaging everything in boxes, use the reusable containers to deliver products to the automated warehouses. Probably massively inefficient in terms of packaging though.
The traditional terms is "lights off factory" or "lights off warehouse" because if there aren't any humans around why bother turning on the lights? You already have some warehouse systems automatically move goods to human pickers who just reach into the bins and put the items there into the box that gets sent to the customer. These AS/RS systems can already look like robotic warrens, I imagine when we don't need humans pickers any more you might see even more density in warehouses with layers of picking arms one on top of the other. You'd still need maintenance access though.
Harder to imagine a society in which we have no need for jobs though
Unless we get something like UBI we'll always need shitty job to keep people busy. Automation for the sake of automation doesn't bring much if 70% of the population is still working 8 hours a day 5 days a week on more and more useless tasks.
This isn't even talking about the fact that we're increasing productivity per unit of time but never reducing working hours, I wonder when this is going to end
> we'll always need shitty job to keep people busy.
That's a pretty darned condescending attitude to harbor there. I know a lot of people who have what you'd call s#### jobs, and they're happy, fulfilled, and proud of the work they do.
Are they ? Amazon warehouse workers ? People working in slaughterhouses ? Starbucks "baristas" working 3 jobs at once while they're execs use every trick in the book to avoid taxes and make literal billions in profit ? Min wage people working insane hours to barely afford life ?
I'm the condescending condescending one ? They're being shat on by capitalism.
I wish only one thing, for their lives to get better. I'm not going to lie and pretend people working 8+ hours a day, doing back breaking repetitive tasks, with no real possibilities of evolving their career are living the dream... all of that to get paid wages that barely let them afford a decent life and never let them escape living pay-check to pay-check
They've been tricked into being proud slaves, that all the pride there is in there. We could do so much better
The condesention is in the term "keep people busy". It is a suggestion that thier jobs are unproductive busy work. They arent. A less condesending phrase might be "keep them employed" or "keep them in work".
This is an extremely black and white way of looking at it, I would suggest you find some grey areas in your thinking.
For what its worth, a high percentage of people are living paycheck to paycheck, no matter how big that paycheck is. Many people will adjust their lifestyle to or above their means, so they have to live paycheck to paycheck to support that lifestyle.
The sheer degree of superiority in your comment is stunning.
I worked lots of jobs before my career in coding. Manual, sometime menial work. The fact that you deem me to be a “proud slave”, too stupid to even see my shackles might be the most stunning condescension I’ve heard, and there’s no shortage of it on the internet.
There is a fulfillment I still miss in that work. I could see the effect of my work immediately, as it was grounded in the physical world. I could leave and the work wouldn’t follow me.
I'm not attacking the people losing their life in these work infernos, I'm talking about the system which enslaved them and made them think this was their paradise
> There is a fulfilment I still miss in that work.
What work are you talking about ? Moving boxes of gadgets all day for lord Bezos while being tracked by optimisation software telling you when to pee or take a smoke break ? Slicing chicken breasts 10 hours a day in a slaughter house ? Driving your leased car for uber 12 hours a day to earn a semi decent wage ? Driving a delivery van, working crazy shifts and not being able to take restroom breaks ?
I'm not talking about a guy making hand made wooden tables in his garage enjoying his craft and skills, I'm not talking about a plumber or a mechanic fixing real world problems, I'm talking about human beings used as robots, doing indecent work for indecent wages.
People aren't stupid, they're lied too, since birth
> Unless we get something like UBI we'll always need shitty job to keep people busy
I find myself often making this assumption too, but I wonder if it's warranted? A "shit" job is often shorthand for a job you can grab someone off the street and more or less have them start working, an "unskilled" job. But the fact that pretty much anyone can do a job isn't the same as saying that someone doing a "shit" job is only capable of that kind of work.
What if we did get to a point where nearly all the tedious, unskilled jobs were done by machines -- what exactly is the barrier to taking all the humans previously doing those jobs and leveling them up to do jobs machines can't do as well? I don't think it's human capability -- I'm quite skeptical of the idea that there's whole classes of people actually _unable_ to learn skills needed for better work. It seems the barrier is that generally it's not economically beneficial to do so -- training takes time and money. But if that's actually the problem, then its not an overall lack of money, it's a distribution question.
Perhaps it's because I'm watching The Expanse now and remember the books, but I'm not completely sold on UBI. I think it gives up on the idea that masses of people can work better jobs. As a supplementary income program I can see it working, but as a replacement for having any job at all I'm not sure I'm on board.
I think the question is what need does further economic development satisfy for the totality of the population.
UBI can ensure no one has an unmet need for basics, but many people with nothing to do would likely lead to unmet fulfillment and security needs. UBI is unlikely to fund digital nomads in the short term, and unbounded travel isn't for everyone. The classic work for independence mindset that drove early settlers also becomes permanently non-viable.
I'm a fan of UBI, but I think it brings us closer to asking the question of how do we structure a society where a 40 hour work week is optional? what will people who are satisfied with the basics do?
Why bother with the higher level jobs, when my basic needs are already met? I think the assumption relies too much on trusting human nature. from personal experience, I’ve known people Who have become complacent after receiving state aid for a while.. obviously, the level of education and culture are strong factors here.
> what will people who are satisfied with the basics do?
Whatever they want, why would we care ?
What would you do if work stopped everyday at 1 PM ? Go to the park, chill with friends, read a book, learn music, go for a walk in the woods, play with your kids, take a nap, go fishing, ... live life ? Are we really so alienated that we can't imagine what we'd do outside of work ?
> What if we did get to a point where nearly all the tedious, unskilled jobs were done by machines -- what exactly is the barrier to taking all the humans previously doing those jobs and leveling them up to do jobs machines can't do as well?
Because history repeats itself. Yesterday's factory workers are now slaves of the service/gig economy. These things have been discussed in basically any economy book since the 60s. Automation in itself just moves the problem, if there are no societal changes to assist it you're just running in circles. Politicians of the 70s were promising 5 hours work days and 3 days of work per week thanks to automation, last time I checked we work as much as back then (if not more because retirement age is going up)
> It seems the barrier is that generally it's not economically beneficial to do so
It's not economical because in today's world you need an army of cheap labor to generate big money for the sharks on top. And this is exactly why you'll always have shitty jobs paying the bare minimum for insane amount of work, it'll take a looooong time before a 7$ an hour worker gets more expensive than a robot for any task remotely difficult.
I'm not saying UBI would solve these issues, or any at all really, it's just something we should at least give a shot. But the problem is that in our societies being a worthy human being = working 8 hours a day 5 days a week, and as long as this is default we'll never get rid of shitty jobs, no matter how much we automate, we'll always have to find new occupations for people, and by definition these jobs will be more and more abstract, less and less stimulating, &c.
> what exactly is the barrier to taking all the humans previously doing those jobs and leveling them up to do jobs machines can't do as well?
One challenge with this that I rarely hear talked about is that many people aren't inclined to do those jobs. Here on HN, we tacitly assume that everyone loves intellectually challenging, mentally complex work, but that's simply not the case for many people. Lots of people prefer to use their hands, or their bodies, or to do jobs that are primarily social. Many (most?) people don't consider thinking to be a deep source of joy and satisfaction in their lives.
When we automate away the skilled manual labor jobs, we implicitly say that the only kind of being human that has value is brain work, but that is an absolute aberration in human history. For most of Homo sapiens evolutionary history, it was just as important to have physical and manual skills.
A fact of life of job-training programs, is that you can't force someone to upskill. (see: coal miners). I think there's generally a strong culture of continuous personal improvement among millenials, and probably late gen-X-ers as well.
But here I am, later in my career, and I don't honestly see the reward of upskilling. For the time, money, and effort it would take. There's not a lot of things I'd rather be doing than what I'm doing now. And some of those things are maybe 5-10 years of training down the road. And then I'd be starting out at a Junior-level in pay, which is completely un-doable in my situation.
Of course, I'm not obsolete in what I'm doing yet. So I don't have THAT choice to make.
> I'm quite skeptical of the idea that there's whole classes of people actually _unable_ to learn skills needed for better work
According to Jordan Peterson, around 10% of the population has an IQ below 83, (at which point the army will no longer take you).
> but as a replacement for having any job at all
I find it unlikely UBI can properly account for the market if it's just a flat rate. Instead a more complex version of "food-stamps" might be better i.e. discount items of food, clothing, accommodation etc with category specific UBI. If people can live on basics without working, so be it - just as long as there is no competition with taxpayers.
That said, there are some things that are still difficult: insurances, medical and crime for example.
> what exactly is the barrier to taking all the humans previously doing those jobs and leveling them up to do jobs machines can't do as well?
Suppose the cool jobs require knowledge of physics, linear algebra, coding, genetics - most people will be unable to do these jobs or unwilling. There are going to be human-in-the-loop kind of jobs for monitoring AI models though.
And there are going to be new applications that will keep us busy, whenever we gain new abilities we also expand our ambitions - like the car industry, which killed the horse and buggy industry but created a huge number of new jobs.
> I'm quite skeptical of the idea that there's whole classes of people actually _unable_ to learn skills needed for better work
Disability recipients fall into both categories. Some are not trainable for skilled work with current methods, while others are trainable but not desirable (lack of current work history, lack of hard/soft skills that only come from work).
In short, they're getting some training but not getting the work from it. That has to change somehow.
> It seems the barrier is that generally it's not economically beneficial to do so -- training takes time and money.
Seeing that a lot with entry level IT. It is more economical to wait for the unicorn than to train the people that are available.
What would have to happen to change that (short of a miraculous repeat of the late 1990s)?
Is this the end game of humanity though ? Let's create more bs jobs so that people can continue to afford new shiny gadgets ?
> ALWAYS
Earth is a finite environment, with finite resources, at some point it won't scale anymore, especially with today's ecological concerns. The world isn't going to run on cells phones and TV shows indefinitely, and even if, I think we should strive for more than that
> People just always want better things and will work for them.
It hasn't always been the case and it won't be forever, capitalism is fairly new, I hope we'll reconcile with our humanity at some point, there is more to life than working 50%+ of our awake time to watch netflix and buy iphones
Nobody said better things require more resources. If anything, modern manufacturing uses far less than it did decades ago, with less pollution, people renting things more than wasteful owning, etc. E.g. when driverless taxis come along, expect to see car manufacturing plummet.
> there is more to life than working 50%+ of our awake time to watch netflix and buy iphones
Life is what you make of it. If you want to write poetry or symphonies in your free time, or watch Netflix and play Candy Crush, I'm not going to judge you either way. You have the choice to make your life what you want. You can focus on yourself and you don't need to judge others.
> If anything, modern manufacturing uses far less than it did decades ago, with less pollution, people renting things more than wasteful owning,
I don't think we live in the same world then, we've never imported/exported so much shit, third world countries are catching up with the west in term of how they want to live (personal cars, bigger houses, more electronics), we're polluting more than ever, &c. all the metrics are going up, not down
> Life is what you make of it.
Well, yeah, exactly, but what can you make of it when you work 8 hours a day 5 days a week for the 40 best years of your life
> You can focus on yourself and you don't need to judge others
Or I can think about myself _and_ others by thinking that the world would be a much better place for everyone including myself if people didn't have to work so much for so little.
Who told you that modern manufacturing produces less pollution? It may produce less of some things, but plastic and evergy use is way up in recent years. For example, a modern car has far less recyclable or recycled content than one from the 80s (more plastic, less steel).
On top of that even if we made cars (and other things) less polluting we still use xxx% more than 50 years ago.
I always use the plane engine analogy, 10 years of research gives you a 3% boost in fuel efficiency, during the same time you doubled the number of flights per day ....
We're already seeing how we badly impacted earth* not even 200 years after starting to pollute it... I'd like to see how we'll make dead rocks without atmospheres more comfortable than home before we collapse
* literal biological paradise we've been engineered for millions of years to live on
But now you’re moving the goalposts a little. If the problem is limited resources, the solar system has a ton of resources, and we can use those resources to improve life on Earth. Eventually we are going to learn how to live in space by managing small habitats there, but as those habitats gradually grow in size and in number we will learn from that and even become better at managing the very large and complex habitat that is the Earth.
Guaranteeing employment by creating bullshit jobs is one of the big mistakes made by the late USSR. Doing so represented a fundamental shift in ideals from the egalitarian "from each according to his ability, to each according to his needs" to the puritanical "the who does not work, neither shall he eat". I keep thinking recently that the US looks more and more like the old jokes about the USSR.
There is no fundamental reason that you need to "keep people busy" with unproductive labor. Due to increases of productivity, we could all be working 20 hour weeks today and enjoying the same standard of living as 50 years ago.
Dont worry we will all eventually just be working for the military-industrial complex in some form eventually, directly or indirectly. That’s basically the end state.
I think there are some jobs where the role of the job is to - for lack of time to come up with a better phrase: "help humans struggle with human things".
Yes, a nurse's role is to heal, but also to offer comfort to a patient, a teacher's job is to instruct, but also to try different ways of connecting with a student struggling to understand an idea.
I don't know if people think those types of jobs are bad or not. I think some of my family members would have had better hospital experiences if they could have had more attention from staff, so I don't think I'd call it useless.
I think that automation could lead to a world where more of the humans-dealing-with-human-things types of jobs are being done.
They do exist already, sort of. Fully automated pallet racks and automated warehouses for small articles do exist. They are very capital intensive so and scale not really good.
I feel like, for a long while, you'd still need people to clean the place, to maintain the robots, install new equipment. There's also going to be occasional messes. Box was badly taped, falls apart, shampoo bottles get crushed on the floor, etc. Also, if you have a robot that puts things in boxes, you need someone to restock the boxes, unless you have another special-purpose robot just for that too.
I'm sure we can eventually have truly general-purpose robots, but in the mean time, I'm kind of skeptical about how much cost and labor these things are going to save. Not to say this isn't impressive or cool, just saying that a completely fully automated warehouse is still pretty far away.
Save? Plenty. It means less parking spots, less HR, less toilets, and so on. They don't get ill, don't have good days and bad, they are fully trained the second they start, and so on.
The most difficult part of business is the human element. For better or worse these mitigate that. The resources freed up from addressing the human element can focus on other opportunities.
It's also questionable how much cheaper robot training is. Specialized programmers to write special-purpose code just for your specific application don't exactly come cheap... Especially when you're not paying them directly, but rather trough some external robotics company you're consulting with. We could easily be talking $300/hour.
Robots dont get ill? Talk to anyone who maintains complex moving machinery such as aircraft or industrial machines. They have bad days. They have lifespans and old age. Some develop long-term diseases (corrosion). Some even get viruses. They can be down for months, even years, as such faults are corrected.
When was the last time a robot took an employer to court for a slip and fall. Most (warehouse) employers would rather have one FT employee doing robot repairs, than a warehouse filled with humans.
Yes, of course robots aren't indestructible or immortal. But they're still easier to manage than the majority of humans.
I don’t know. This doesn’t seem dramatically better than the Unimate industrial robot, which is over half a century old. It even used the same kind of suction tool:
https://youtu.be/VdolSBpyCaU
I mean, Boston Dynamics’ new robot is on (Mecanum?) wheels and it has some basic machine vision capability, but it really doesn’t seem like a dramatic, exponential departure from the 60 year old Unimate robot.
(Boston Dynamics’ other robots are much more impressive from a technical standpoint, but I’m not terribly surprised a company that has been struggling to get profitable is turning to more mundane but practical robotics applications.)
Which doesn’t mean I don’t think something like this isn’t practical. It just goes to show there’s still a ton of fairly virgin ground out there for increased automation and productivity growth just deploying technology from half a century ago and adding some minor tweaks. And I think part of the reason we HADN’T deployed this kind of automation tech earlier is because wages have just been too low, with wage growth fairly stagnant after the 1960s.
Every day I’m less and less worried that automation will get rid of all our jobs and more and more frustrated at the slow uptake of productivity improvements we have just sitting in front of us, perhaps because companies don’t compensate their employees enough for it to be worthwhile.
I don't think Amazon cares much about the long team warehouse worker situation because they know it will be automated very soon. I'm curious if workers unionizing could negatively affect their automation aspirations.
They certainly act as if they care. They should automate as these shit jobs shouldn't exist, and if workplace is organizing ensures the automation becomes economical, that's a good thing.
Well, they care about their image, I just don't think they care much about this social issue. According to them, people that have to pee in bottles would never work for Amazon.
The Boston Dynamics PR machine is in full swing. A recent acquisition, the 60 minutes interview last weekend, this recent press release... Seems like they need to figure out how to make some revenues.
The problem is they based all their tech on old-fashioned control theory, which will have its ass kicked by deep-learning-based tech when the first serious competitor comes along.
Doubtful. You need provable security and systems characteristics if you wanted to get approval for a household robot in for example the E.U.
Higher cognition and task solving? Sure, we can do that with blackbox methods, but everything where skulls could get crushed, you want, and by law need, the boring stuff.
You can reverse the reasoning: if you can't safely manuoever a car through a street with people, then how do you expect to do the same with a legged robot?
I'm not saying that machine learning isn't capable of doing these things (most of the time). I'm saying there's no way in hell you'll get a non compliant blackbox system past European safety laws.
But robots need to do more than make accurate movements. They need to detect where they can safely move. For this last part, you can't realistically make provably correct control algorithms (unless you disallow interaction with humans). Hence, you might as well use deep learning for the entire system. Also, you can limit the damage the system can make in some other way (e.g. limit the acceleration of the robot or make it lightweight, etc.)
Right on time for Amazon's union vote!! I support people's right to organize, but warehouse workers are in for a rude awakening. Amazon's warehouse "partners" are just like Uber's drivers: intended as stopgap labor before AI can stock shelves/pick items/drive cars; except that stocking a warehouse/picking items is orders of magnitude easier to accomplish than autonomous driving. I look to Amazon to start shedding jobs by the tens of thousand in the near future.
The more automation happens, the more desperately cheap labor there is to disincentive further automation. Organizing to raise workplace standards is in fact necessary to ensure the shit jobs are entirely eliminated.
I really like how the press release talks about multiple value props instead of the robotics tech. As a techie business owner who spends most of my time in sales this is something I need to constantly remind myself about. People want the "why" not the "how".
We are rapidly approaching the automation of most repetitive human labor. Warehouse workers, truck/uber/cab drivers. Fast food workers, etc. What happens when these jobs go away? Historically technology has always provided a means of replacing those lost jobs but I think we are reaching an event horizon now where all of that changes. What happens to all of the people whose positions are replaced? I think we are teetering on the brink of a need for UBI, a massive change the world style project (populate another planet or something on that scale) to keep people engaged, WW3, or mass unrest on a scale we have not seen before. Or perhaps and most likely we just ignore the newly poor and homeless and those of us that can continue to accumulate things and tssk tssk at the people that wont pull themselves up by their bootstraps.
This is how technology always works. 200 years ago, most of us worked on farms. 100 years ago, most worked as domestic workers or in factories. Now most are white-collar or service workers. As technology evolves, so too do our jobs.
The problem: it's true that productivity per person has steadily gone up through technology, but there are still a fuckton of workers in farms, mines, and factories to make commodities for the developed world to consume voraciously. It's only that these factories are mainly in the third world, and most people in Western societies can't see this. Even with the purpoted advancement of technology, miners in Congo are still mining cobalt with shovels (https://www.youtube.com/watch?v=fPIB17PE2vM), Columbian farmers are still picking coffee fruits by hand (https://www.youtube.com/watch?v=VNJ2hU3dr8o), and Chinese workers are assembling iPhones by hand (https://www.youtube.com/watch?v=TmLsV9cSk0o). These kinds of work, even though you may find "primitive" and "solvable by technology", will take a long time to disappear off the face of the earth. This deterministic "evolutionary" thinking of jobs is really an incredibly narrow take on the world economy of jobs as a whole.
As most of us here are in the technological sector we are often blinded by the idea of unending linear progress, but we really have to respect how hard robotics actually is. The current manual labor tasks that have to be automated are optimized are actually not that 'repetitive' in the scientific sense, they actually require a lot of elaborate limb movements, split-second decision makings, and handling of countless edge cases for even just the simple task of moving objects from point A to B. It's easy to manage this when you're in a warehouse that has been purposefully designed for robots to do things easily (with entirely flat floors and standardized box sizes). But as soon these robots meet environments where humans live, come and go, with all the irregularities and edge cases it has, it will fail miserably. Too much of our surrounding infrastructure are designed for humans and not for robots, that's why there is immense value in creating robotic systems mimicking natural systems. But robotics researchers are still having an incredibly hard time replicating the delicate movements of human limbs and hands (both in terms of hardware and software), and rightly so; the human body is an incredibly elaborate system of muscles, bones, and neural pathways that almost seems like a miracle of nature (how the fuck did evolution really create this insane complicated system that can do all sorts of weird movements???) And even if we know how the whole system works, it's not a easy task to replicate this in mechanical terms. Note that we haven't begin talking about mass production of these complicated robots, there's always a question of whether it will scale (and would it be more competitive to humans, especially in the developing world?). I think robotics will still have quite some progress in the 21st century but a bit pessimistic that it will seriously begin to replace traditional manual labor in my lifetime.
It's a mobile palletizing robot. Palletizing robots have been around for years, but not, usually, on a mobile base.[1][2]. Fanuc (Japan) sells large numbers of robots for this. Handling uniform cardboard boxes is routine. Systems which handle a larger variety of items are rarer, but do exist.
Automated truck loading and unloading robots are already available. It's neat that Boston Dynamics made a mobile one, but they are not the first to do that.[3]
The real frontier in this area is fully automated order picking of miscellaneous objects in fulfillment centers. Nobody has that in volume use yet, although there are many demos.
reply