Hacker News new | comments | show | ask | jobs | submit login
Walmart will soon have robots roaming the aisles in 50 stores (businessinsider.com)
354 points by hourislate 11 months ago | hide | past | web | favorite | 316 comments

The immediate take away i get from this is that it really isn't about taking over human jobs. humans don't do this job nearly as frequently as the robots would do it, because in a standard retail environment like wal-mart, it doesn't really matter all that much if the reported in-stock quantity is slightly off for an item. that is to say, it matters, but not so much that the big-box retail environments i've worked in would dedicate more than one full-time staff to do this. if the robot is working 24hrs, it's doing the hours of three full-time staff and probably the work of six. That's a huge jump in the amount of stock checking that's happening at a store.

So i wonder why wal-mart is investing in robots to fill this role, when there are so many other important jobs that they could automate instead. and the answer i land on is that the quantity-on-hand is critically important if you're using the store as a warehouse for a local delivery operation. If you've advertised something as in-stock for same-day delivery on your website, and then you let customers in to browse around your warehouse and move things around, having roaming scanners checking for mis-homed items is suddenly a lot more important.

Or it's an experiment. They're giving it a lower importance job during the pilot program. Hard to imagine the initial deployment being on a mission critical path for the store's operation.

It's probably because stock checking is a simple repetitive task that humans tend to fail at due to boredom and mental fatigue while robots would accel at it because of the fact that it's simple, well defined and boring.

It's traditional to spend 45-180 seconds in the back room, staring at your phone or talking to a coworker, before returning to the sales floor to tell the customer that you're sold out. It's like the Hello World of stocking product.

In my experience about a 10-15 years ago when I worked in a super-market there was no "back room" really - it was just a staging-post for getting stuff off the truck, putting it onto trolleys, then taking it out onto the shop floor to put on the shelves as quickly as possible.

I dunno if people expect there to be this huge warehouse of stock "out the back" of every supermarket with loads of shelves holding tins of beans or loaves of bread or whatever just sitting there. The shop floor was the "back room"!

I used to just tell folks that what they see is everything we had, so if they cant see it then we dont have it.

Even with fancy modelling and "just-in-time" delivery stock-outs still happen - I remember one year working on the last weekend before Christmas. We "sold out" of turkeys - in reality it was just that people had them in their trolleys and hadn't checked-out yet so people were stealing them out of other customer's trolleys when they weren't looking! Happy Xmas! :-)

When I worked in a supermarket in the late 90's, we did have a pretty expansive back room filled with stuff that had been taken off trucks but not stocked on shelves. However, it was next to impossible for a front-end person to actually get to any of the products - they were stacked on pallets 3 to 5 rows high and about 5 rows deep with no clearance between them. Occasionally someone would ask me if something was in the back, but unless it was in the dairy fridge, or a case of it was actually sitting on a cart in the hallway, I'd tell them I couldn't find it.

The exception to this is milk and bread. There’s always more bread and milk in the back.

Or somewhere else. Some stores have the same item at multiple locations which one may not know.

Ahh yes.... "Stealing turkeys out of fellow customers trolleys", the TRUE spirit of Christmas.

It only checks inventory at this point because there are no robots that exist that are capable of doing the rest of it safely and effectively such as stocking the shelves in those uncontrolled dynamic environments.

When they get more capable robots, they will replace more jobs.

Once robots can fix the shelving problems cheaper than humans then they will replace those jobs.

As it stands total cost to fix an individual shelving problem likely dropped. In this case it'd be surprising if Walmart employed more people to fix shelving problems, but in other contexts we may have seen induced demand for more employees because the ROI for employee time improved.

Not all increases in efficiency reduce the # of total jobs, we should expect to see the total # of jobs increase in some circumstances.

And the workers are now focusing more on processes that could be argued are more skilled, like restocking the shelves, fixing prices, etc., than they were doing before, like identifying issues.

They are trying to increase the productivity of the stock crew when the store is closed, enabling them to shed a couple of relatively expensive overnight stockers per store. These robots won't be operated during business hours, but take over or make easier some of those stocker duties.

The robot will scan the shelves in a pattern, followed by a stocker with a cart to collect all the misplaced items. It will print shelf labels, without the need for a stocker to use a cumbersome printer relatively slowly.

It also to me seems it might be a way for them to in-house inventory control services like RGIS, using robots to replace human counters. If it can detect misplaced items, one day it might be able to count placed items.

It could also be for the same reasons you stated: it's low-controversy and low-responsibility.

Low-controversy, but visible. Wich will introduce people to the idea gently. They can add features to the robots later with less resistance.

availability is a psychologic factor in purchase decisions. The more in the soup pot, the more soup people will eat. Whereas the beloved low stock fear of missing out trigger might be too hectic for retail. The warehousing aspect is interesting too.

Also consider that better inventory data allows better just-in-time inventory refill, increasing $/ft^2 inventory turnover and therefore total $ sales per store. I suspect this is about driving analytics with higher quality, more frequent data.

Yes, I think this is more likely what they are doing. With people shopping the store, they can never really be sure about on-hand inventory even with robot checks. The robot could confirm that an item is on the shelf, and then a moment later a customer could put that item in a shopping cart and not check out for another 30 minutes or more. During that time the on-hand count is wrong.

Stock levels are important for home delivery but this robot is also a logical first step for a fully fledged shelf packing robot.

If the current robot signals something is out of stock but a human determines the robot is incorrect then it means the store map is wrong or the product identification needs work. Once the error rate is low enough then full shelf packing can be implemented.

Have you ever worked in a store like Walmart? Robots won't be stocking shelves without major(!!!) changes to how things are displayed.

Considering I've actually called stores to check whether they have a particular item in stock or not, I imagine you are not too far off in your hypothesis.

in the distribution system I am in, excess inventory is concern. not only identifying over stocked items but rounding up all obsolete inventory taking space in stores and warehouses. then understanding why it is there.

since wal mart doesn't close stores often or deal with private stores I am curious how they account for take backs from such. that alone can upset the inventory of any business.

accurate inventory is so key with both the seller and supplier. I am not surprised they want to focus there. plus another issue is shrink, by having robots fulfill the role the chance of it vanishing on a count is gone

Relatively more real-time big data, AI/ML or at least sophisticated algorithms to detect or even predict shoplifting.

Stock checking seems vastly easier to accomplish electronically (a la RFID tags) than with dedicated robots.

Porque no los dos?

As other jobs get automated, inventory checking could get more important.

The answer is that this is one of few jobs that is simple enough that it can be automated.

> The robots are designed to free up store employees' time so they can use it to help customers.

...which is bullshit; Walmart just doesn't want the "robots taking our jobs" angle here. If they consider the current level at which employees help customers in the aisles sufficient, then they'll be cutting jobs to match that level rather than "freeing up" employees to do more of it.

On the other hand, Walmart is certainly feeling the pressure from Amazon and improving customer interaction is likely high on their strategic radar. To wit: AMZN revenue +33% to a $170 Billion run rate, vs. WMT revenue virtually unchanged over the last 4 years at $485 Billion. Walmart recently introduced their own version of Prime, and AMZN is now in the grocery business, two more differentiators that have been closed off between the two.

It doesn't take too many 30% years to go from $170 to $485 Billion (not that they'll continue at that pace, but shit, this is year 23 for AMZN, and their growth is still incredible, [AWS, AWS, AWS]).

For the foreseable future, robots and chatbots will be poorer UX than speaking to a human. Since it's not adding value to their product it's obviously being done for cost cutting and scale reasons.

I rarely get useful assistance in large chain stores. Asking "where is the pesto?" usually gets an answer like "I dunno, somewhere in aisle 7 maybe" with an indifferent shrug. I'd happily make three attempts at using a clunky voice interface if I got the answer "Aisle 7, bay 11, shelves 3 and 4. We have 18 different kinds of pesto currently in stock. Would you like me to take you there?"

It almost seems too obvious to say, but computers have different strengths and weaknesses to humans.

There are also social factors involved that may be advantageous to machines. Self-checkout machines offer an objectively worse experience most of the time, but a lot of shoppers prefer them. Some customers don't want to interact with a human cashier. Some will be buying embarrassing items and prefer the apparent anonymity of self-checkout. There's also a subtlety in how the self-checkout machines are arranged - because there's usually a common queue for several machines, customers don't feel pressured to finish their transaction quickly.

Shoppers may prefer being assisted by a robot for a variety of reasons, even if the robot isn't quite as good as the average human worker.

I already choose grocery stores with ample self checkout, which is so much better to me than waiting in line then interacting with a clerk. I for one welcome our robot service providers.

Interesting - after the "new shiny toy" factor wore off, I've deliberately gone through the staffed checkouts where at all reasonable to do so. This way I feel like I'm keeping someone's job relevant and maybe even keeping them employed.

I admit this isn't a very rational response, though, because I'm fine with self-service airport check-ins and generally pretty happy to talk to a machine rather than a human.

I use staffed checkouts for a different reason -- it is vastly faster if you are purchasing more than a few items.

All I have to do in a staffed checkout is remove items from my cart and place them on the belt. The cashier scans them, and can begin this task while I'm still emptying the cart. And most of the time, the cashier is practiced at this task and can do it quickly. When I'm in a self-checkout I see seconds wasted per item as people turn them over and around one-by-one to find where the bar code is.

And the staffed checkout usually has another person to bag the items, and again can begin doing this while the cashier scans more items, collects payment, etc.

At a self-checkout I'm doing all these tasks myself, stepwise because the computer insists that I bag each item before I scan the next one.

I always choose a cashier because they are, without fail, more efficient at getting me checked out than having me fumble with a checkout machine. And that's without having any sort of fresh produce. If I have fresh produce then the machine is even worse. Making me do the job you once paid someone to do? No thank you, grocery/bigbox store. Maybe once in the last 10 years when I had a single item and it was holiday season -- even there express checkout is usually faster.

It does raise the question of what jobs are worth saving? So a robot takes someones job, but it was a crappy job that they hated anyway. But they needed the money. Tough.

That is a generic end-of-work question. Ideally, everyone would have jobs they liked, and all the dull stuff would be automated (Star Trek outcome). Pessimistically, the machines will become better than humans at everything and we will just become obsolete (Terminator outcome). If automation picks up, we will have to deal with these questions sooner rather than later, but it will happen eventually.

I have a friend who is happy to do the dull work. Smart guy, just content. At first I encouraged him to move up, then I began to understand that not everyone views the world the same way that I do.

Let's be optimistic and hope for the Wall-E outcome instead of the Terminator outcome. The robots will keep us around and take care of us as pets.

There's a middle ground portrayed in the Culture Series novels. Everything is run by machines for the most part, and in some situations humans just tag along for the ride. At the same time though humans can basically get whatever they want whenever they want it, and do whatever.

Erm, if we create robots better than us, they are our legacy, our homo superious, why should Homo sapiens continue? You don't see Neanderthals running around today for a good reason.

It’s easy to look back fondly on genocides that have already happened. No one shed a tear for the long lost Neanderthals. The Estruscans were wiped out by the Romans with almost no recorded history left behind. Native Americans, First Nations, and Aborigines suffered horribly at European hands, but we never entertain the idea of moving back to the Old World.

Things are different for genocide that hasn’t happened yet. If people have left the UK and formed the United States which is far more successful currently, why not just kill all the citizens of the UK? Obviously the US is homo superious.

Would no one shed a tear?

It wasn’t necessarily genocide in the Neanderthal case, just basic species out competition. Yes, many people today are trying to “save the panda” even though it is quite non adaptable, but if the pandas go extinct because they cannot compete for the resources they once monopolized, that isn’t genocide.

At least in my mind, genocide is unnatural.

Playing devil's advocate: Is there actually a difference between genocide and survival of the fittest? If the Etruscan people were a superior culture with a superior fighting force, they would have defended themselves from the Romans and they'd still be around. Likewise if the panda was better adaptable, we wouldn't be struggling to save the species.

If the saber-toothed tiger died out because they were out-competed by the other species around them, we call that natural selection. If the Carthaginians died out because their neighbors had a better fighting force, we call that genocide. But the only difference is... what, exactly?

It's not a linear outcome space. There's also the outcome where automation enables tons of new dull (but in different ways) jobs, currently embodied by the so-called gig economy. I think this area is nacent and has a lot of potential[1].

Also, incidentally, this outcome is consistent with a couple of centuries of economic history.

1: No, these jobs aren't great compared to classic middle class jobs. But going from shelf stocker or check out clerk to Instacart shopper and/or Uber driver isn't obviously a step backwards on net.

I hope for the Culture outcome. The machine minds gets better at everything, and they choose to keep us around as pets. (Well tending to all of our Mashlow levels, as a point of pride between themselves.)

Skynet is just sloppy though. The easiest way to get a practical human extinction is to make entertainment so fun that people forget to have children, then just wait.

Its kind of ridiculous to paint the future as binary. Computer augmented humans, better than robots, are also a clear possibility.

I can't agree with this enough. Taking it one step further: I actively _avoid_ grocery stories that don't have ample self-checkout. If I have to deal with a human clerk once, I'm way less likely to return to that store.

Trader Joe's is excellent about this. I'm not sure if it's explicit policy but every single time I've asked an employee where something is they dropped what they were doing, walked over to the exact spot and literally handed me the item I was looking for. I guess that's only possible in a small store though.

This is absolutely standard in the UK in all the supermarkets I'm aware of. If a member of staff doesn't know where the item is, they'll go and find someone who does.

I've found that too. TBH I just want them to tell me which aisle it is, and will ask "which aisle are eggs on, please?" or whatever. But they insist on walking you there and done are seemingly instructed to hand you a product too (which usually I replace).

Mind you, if they didn't insist on moving everything to increase customer circulation ...

All grocery stores I have been to are like that. They will drop what they will tell you where to find it off the top of their heads and offer to take you to it. Publix, Piggly Wiggly, Winn Dixie. This is why I still do most of my grocery shopping in grocery stores and not big box.

Wegmans employees do the same thing, and my Wegmans is bigger than some Walmarts.

Walmart used to do this, back in Sam Walton's day. (I grew up and live in Southwest Missouri, and remember Walmart as our small, regional competitor to the big national giants like Kmart :)

Many stores do this. Our local grocery store, QFC, frequently does it, Home Depot does it, etc.

I have the opposite experience. Pretty much any store around here, Walmart included, the employees are very willing to help you find stuff, and if they don't know where something is, they will find or call to an employee that does know.

I am very impressed with folks at home improvement stores. I walk up, ask them where X is and they say 'ahhh, it should be over in aisle 17, towards the end past the Ys.'

Home depot lists aisle and bin numbers of inventory on their website once you've selected a local store as you're default store. After asking an employee and them pulling out their phone to check online for a products location, I've never had to ask an employee again. Surprised they don't advertise this on in store signage

Item location could be provided by a phone app today. My feeling is that they don't do it because they don't want you to know the exact location because they want you to browse the aisles in hopes that you'll throw something into your cart that you weren't planning to. Same reason that supermarkets put milk and eggs at the farthest corner away from the door.

The Home Depot app is how I find everything when I go there. I think the other stores didn't do it yet because the isles and bays in most stores don't have visible numbers. And without that they need to wait a few years for deployed internal location technology to improve a bit.

Walmart actually used to list exact shelf location for items on their site. Maybe they still do, maybe desktop only?

I've been in a retailer hackathon. This is true. (ratailer is a worldwide actor french born)

An aisle number is all I'm really looking for when I ask imo. I can scan a shelf about as fast as I can walk.

But if I’m asking, it’s because I already looked on the shelf and I didn’t find the thing after looking for at least five minutes. I want to know, at that point, whether they carry the thing at all; whether it’s just in a crazy place (e.g. honey in the bakery in a little spreads display, instead of in the baking or condiment aisle); or whether my eyes are failing to do their jobs properly and I’ll need to look again.

I can't find it if I'm looking at it. I long to photograph the shelf and then search it with OCR.

Part of this has to do with demographics. We prefer to deal with robots and order stuff on web interfaces. I was talking with my aunt and uncle the other day. They were telling me that they prefer to bring their paper checks to the bank and interact with the teller rather than scan it in the app from the comfort of their home.

That blew my mind.

I think there's a huge age gap when it comes to preference for human vs mechanical interface. Until the older generation dies out these sort of redundant service jobs will exist. As they say, sometimes progress only happens one funeral at a time.

It's not really human vs mechanical in the teller example despite what they say. They are just keeping their long running habit rather than having to learn how to scan it on their phone which is work and risking that their money will disappear.

Keep in mind this requires a smartphone. Some seniors don't have them. They may not even have dumb cell phones. Some don't use computers at all. Some are using the old CRT TV they bought ten years ago because it still works fine.

There's a point where your life works fine and you don't particularly need to change it or learn radically new things unless you have no other option at all. In some ways, the old actions have their own benefits. The senior goes to cash her paper check at the bank because it gets her out of the house and in front of people; she saves the money on a cell phone that she doesn't use much because she stays home more than travels at her age.

Even with computers...she doesn't feel helpless or at the mercy of technology she doesn't understand. She might even reduce her vulnerability some that way; a little knowledge can be a dangerous thing.

I think you could easily be wrong here.

I can self-checkout, or at some supermarkets self-scan (using a handheld scanner), but I chose the human checkout operator. It's partly economic, the supermarket doesn't discount my shop if I do the checkout work myself, so I extract a modicum of their takings towards wages for a local person rather than have it go to shareholder profit. People here need jobs.

People do need jobs. I wasn't talking about grocery stores. It was specifically the behavior related to depositing checks.

Also there is a difference between a rationalization for a behavior and the actual behavior. Usually the rationale is actually created afterwards. Although certainly some people must have created a deliberate policy to support employment or simply out of preference for human interaction. But my supposition was just about the check depositing.

Regardless, keeping people in menial jobs is not optimal for them or for consumers and almost certainly not a policy that will generally persist when more capable AIs are available.

>Regardless, keeping people in menial jobs is not optimal for them or for consumers and almost certainly not a policy that will generally persist when more capable AIs are available. //

"Keeping" people in menial jobs isn't the issue. When the robot takes over it's because the owner-capitalists can make more profit by cutting the proles out of the loop.

It's not the job, it's money to pay for water+food+shelter that's the issue.

If rich capitalists are happy to give up their profits to make an equal society with few menial roles then your way s works. There's no evidence that's the case and lots of evidence against it.

Robots aren't going to replace all of the service jobs, but I think they're going to replace more than we might imagine.

Over the last 30 years, more than half of bank branches in the UK have closed. Many older customers have been substantially inconvenienced by these closures, but they're on the wrong side of economics. Beyond a certain point, it's just no longer financially viable to keep a branch open. A lot of older people prefer DVDs to online streaming, but good luck finding a rental store in most towns.

The social effects may well be very ugly, but I think that most businesses will be quite ruthless about their deployment of automation. They clearly have no qualms about the impact on jobs and I don't think that they'll care if some proportion of their customer base is marginalized.

There's good reason to prefer a disc over a stream, especially if you use player software that lets you skip the unskippable previews. The audio is better, the bitrate is higher, and you can rip the disc and store it somewhere safe.

I've almost always had the opposite response, ranging from briefly informative directions to someone guiding me all the way to the item I seek.

(Edit: But I don't shop at Walmart...)

> Would you like me to take you there?

You see another robot going in the direction of aisle 7, but forget it immediately as unimportant. While following the robot something catches your attention (slight disorder in neatly ordered merchandise? You don't reflect on that because...) you see a thing you didn't bothered to search for, but... To the robot: "Organic food is healthier, right? Let's go."

Yes, I think the social factors are underestimated. I actually believe that most people by default would prefer to interact with a robot rather than with another human. Humans have different speeds, moods, knowledge and expertise. When you shop, you dont do it to have a social experience, you do it to get a very specific job done and in my opinion this makes humans a variable that is desireable to be removed.

Walmart employees used to be trained to always escort the customer when they ask for help finding something.

That's if you can first locate a warm body in your geographical vicinity in the first place.

I wonder if we could use a dewey decimal system style lookup service for stores?

What's worse is that a lot of Walmart employees don't speak English. They can barely understand what you are saying, much less, conjure up a response. Usually they have to go and fetch a much smarter or knowledgeable person to help out.

Terribly inefficient.

So Walmart Employees, if this is you:

I'm sorry but if you want to keep working at Walmart, time to step up your game, learn some English and be more responsive. Otherwise, I'll happily accept a response from the robot over your indifference if it helps me get to the product I need to find quicker.

You're being junked because most of the folks in the US haven't experienced what you're referring to. Here in the US everyone speaks English. I understand it's a little different in Canada because there are many more immigrants working at grocers and stores like Walmart that may not have mastered English yet.

I think it's more because "smarter and more knowledgeable" is being conflated with "speaks English." That kind of casual ignorance doesn't elevate the discussion here.

So someone speaking English is "a much smarter or knowledgeable person" than someone who may have immigrated recently. Yeah, makes perfect sense.

> I'm sorry but if you want to keep working at Walmart, time to step up your game, learn some English and be more responsive.

I'm sorry but it's not your place to threaten Walmart employees. You'll have to leave that to Walmart management.

What Walmart do you go to? The one in my town, everyone seems to speak English fluently

Calgary Alberta.

Any Walmart in town.

What's actually happening is a combination of the fact that that they are probably in the middle of stocking and/or immediately detect your racism and arrogance and inability to get to the point and realize they need a supervisor to deal with you.

Sure, you think that, but the last two times I went to Walmart and needed help finding something, the first person I went to at each store didn't speak any English at all. This was in the middle of Tennessee, not Texas or Miami.

Keeping the prices updated and correctly tagged adds value to the product. Walmart in particular seems (from my observation relative to other stores in my town) to have a bad track record of updating their stickers and keeping their shelves sorted.

Ya, they're probably desprate to find a way to compete with Amazon right now. AMZN's been killing them on price and convenience, afaik.

AKA perfectly reasonable economic reasons.

I love the marketing angle.

Just two days ago, our internal company newsletter announced a purchase of another industrial robot, and the newsletter said it is meant to "help/relieve our employees with doing monotonous tasks". In the office, we all read this and were "yeah, right; what you actually mean is 'let us get rid of those employees who are currently hired for doing those monotonous tasks'...".

And that's a good thing. We should be relieving humans of all kinds of monotonous jobs, which means that yes, we'll not have those jobs, and people won't be able to get this work; and again, that's a good thing - a job isn't some kind of gift or value by itself; if some task isn't intrinsically fulfilling for humans, then we should be working to get humanity rid of such tasks.

Coincidentally, "tasks that aren't intrinsically fulfilling for humans" is pretty much synonymous for work, if it would be fulfilling by itself then quite likely that it'd be or become an unpaid hobby instead of a paid job.

I strongly agree, but that's a conditional agreement - I believe we should relieve humanity of the very concept of earning a living, but at the same time we also need to do it in such a way that the whole civilization won't collapse in the proces.

My only complaint in the previous comment is about the PR double-speak. We all know that when a company purchases automation, it's meant to replace human workers. But companies won't honestly admit it.

Prof. Richard Wolff proposed solution to this is worker-owned cooperatives. You'll see a lot of his work on this on his books and in interviews and talks.

The Mondragon umbrella of cooperatives in Spain, a multi-billion dollar corporation had to cut down employees during the recession. But instead of simply firing people, they were given options including offers to re-skill and compensation (like commute). The whole corporation paid for this. Even in this ultra hostile capitalist world, we've now a case where a large scale cooperative can survive the cycle of capitalistism's overproduction and over-exploitation.

Ofc this is all work in progress, and nobody knows what problems it will face. But certainly we need to do something about the current Capitalism.

exactly my thoughts: automation is great, but I miss discussion about what to do with people who lost their jobs. And in the long-term, more and more automation will make more and more people unemployed so social aspect is really important here so if we want to avoid social collapse we need to start talking about it soon.

This is good for bitcon

Or they're doing the "harbor freight" method of having ~3 employees in the store at all times. When there isn't a rush at checkout you actually get good service because under-staffing means people don't specialize so every employee probably knows the answer to your question.

> which is bullshit; Walmart just doesn't want the "robots taking our jobs" angle here

Some would say the "robots taking our jobs" angle is also bullshit.

Damn tractors/threshers/combines, taking all of our jobs....

In UK the initial industrialisation of farming forced people to migrate to the town's to seek work. Their standard of living declined considerably; great news for the mine & mill owners though as they could grow richer and fatter with no dearth of cheap labour, particularly children.


Living standards did rise eventually, but there was a considerable lag. Initial income rises were at costs of health and of hard work, at a pace dictated by the master, for all the family.

I fear we'll see very similar patterns again.

> the robots are 50% more efficient than a human doing the same task. They can also scan shelves three times more quickly and are a lot more accurate. Human employees can only scan shelves about twice a week.

Personally, I want to see up-to-date retail stock mapping (that I can search)

Walmart is testing a robot from Bossa Nova (Pittsburgh and SF)



> Personally, I want to see up-to-date retail stock mapping (that I can search)

Just got from Lowes and they do it on the web and in their app. Search for an item, and if it is in stock it will show you its location on the store map.

OTOH, every Walmart I've visted is unable to keep their shelves full. [1] How hard it is to keep something non-perishable (like deodorant) in stock? Somehow Target figured that out.

[1] https://consumerist.com/2013/04/05/walmart-employees-tell-co...

Anecdote time!

I bought a water softener at HomeDepot about three weeks ago. Initially went to my local Lowes where I had checked if one was in stock and at the store. Yes and Yes. Get to Lowes and there is no water softener, as the website indicated. Lowe's employee checked the back as well, wasn't there.

There are many problems with humans doing this "inventory business" at $10 an hour or so. The professional counts that happen once or so a year by outside agencies wouldn't be necessary otherwise.

To be fair, I've had issues at Home Depot as well... lots of; "The computer says there are 2 in stock... I have no idea where they are."

You’d think it’d be a simple matter of 1. automatically decrementing a counter when the item is scanned at checkout; and 2. manually decrementing the same counter when you throw things out for breakage or expiry (i.e. pretend the dumpster is a customer and ring up its purchases.)

That would just leave theft, which is usually predictable with ML (this being part of what loyalty cards are for), and doesn’t happen at all for most low/medium-ticket items.

>doesn’t happen at all for most low/medium-ticket items.

Everything gets stolen. Mostly by employees. When I worked retail somebody got away with a treadmill. Just wheeled out the door on a cart, nobody said a thing. Later I was looking for a treadmill after a customer had checked the online stock count, and the system was off by three, not just the one I knew had disappeared.

The other big problem with the counts is just stuff moving around the store. It's a limited physical space, and it isn't all merchandised according to the book. Un-recorded promo displays get set up, things get moved onto top shelves and departments where they don't belong.

I was checking online to verify local stock at Lowes and Home Depot last week. Both websites stated '2' in stock. I had to stop in both stores BC Lowes diddn't have another item I needed, so being curious I checked both stock counts of the item. Both had more than 10, which made me think: if item count >= 2, display 2

Obviously it's not the only explanation, but I assume they use marketing research to help decide what inventory counts to display

Yeah, it isn't perfect. Sometimes I see more variety on the shelves then was reported by the search results. But my primary use case is the store map. It is so easy to get lost in Lowes/Home Depot.

Home Depot's website can actually show you where items are on a map. I discovered this because my local Home Depot used to be a KMart and has the most bizarre layout I've ever seen.

It's too bad Home Depot doesn't have wifi and LTE reception is often crappy inside their big box. This is amazing when it works, but so frustrating most of the time when you can't actually load the info without going outside.

There are software solutions for managing resupply, even for perishable products like fresh meat where you want to minimise write-offs: https://www.blue-yonder.com/en/solutions/replenishment-optim...

> Personally, I want to see up-to-date retail stock mapping (that I can search)

Isn’t something Walmart doesn’t want you to have? Having a mapping means being more efficient when shopping in the store, which means less time in it and so less opportunities to buy things.

This reminds me of the argument some (now dead) portal page/search engine company made to Larry Page when Larry and Sergey showed off their page rank algorithm.

It was basically something like the following: "This is no good - if the search results are too good then people won't stay on our page long enough to see the ads"

This is basically what triggered them to start their own company instead of licensing the tech (I'm remembering most of this from In The Plex which I read a while back).

Basically it's stupid not to make the core thing better - especially in the case where Walmart's real competition is Amazon.

Market pressures always lead to quality decay over time. Even with Google - it started with lightweight, quality, bullshit-free search. Look at it now.

Especially Google Maps, I'm waiting for a competitor to blow it out of the water. Satellite images are relatively cheap, you could have a map that shows a daily updated image of every street, and show timelapses. But their streetview would be very tough to displace.

Their mobile maps are bloated and their web maps have been better, but I'm not sure how Google or anybody else can get cheap daily streetview imagery. I've been waiting for somebody to partner with the postal service and slap cameras on every truck.

I think parent was refering to satellite imagery; not street view.

From iOS and Android apps you can search a for products in a specific store. It displays Aisle number and some broad stock information. Maps have been trialed as well. There are challenges with data quality to get maps to work well. Poor data quality is due to managers being able to move shelves around without updating central systems and stock living in multiple places within a store (same SKU). Locating consumers on a map has also been trialed with LED based indoor positioning. Walmart do try lots of stuff and work with lots of vendors. Rolling out to 4000+ stores is a multi-year challenge. See eReceipts, Savings Catcher, Scan'n'Go, Walmart Pay et al. Usually 3 years from pilot to chain.

Walmart's core customer personas are "busy moms" - they know that, in the long term, providing mom with convenience, efficiency, and low prices wins over trying to force her to waste time lost in the store.

>> Personally, I want to see up-to-date retail stock mapping (that I can search)

If you search an item in the Walmart mobile app, it will tell you the aisle that the item is in, in the store you are in (there is an automatically-entered "store mode" that is driven by location detection.)

They have all the data across 5,000 stores and >100,000 SKUs. The robot helps ensure that customers haven't moved items from where they are supposed to be.

> http://www.bossanova.com/

Oh no, it's _ugly_!

I don't think it's that bad. It's meant to be practical and inoffensive rather than especially stylish.

If, at least, it could have been offensive!

They missed a perfect opportunity to give it legs and have it clomp around the aisles muttering "gonk... gonk..."

I see what they did there. Bossanova --> nova boss --> new overlord

Why does it have to be a robot riding around? Couldn't the same thing be done from a camera on the ceiling?

It would probably mean that the camera would either have to hang down pretty low in the isles or the isles would need to be super wide so that the camera could get a good view angle of the entire shelf, including bottom rows. But sure I suppose they probably could have something that rides around on a track on the ceiling and raises and lowers to get into position in each isle.

Like Oto (sp?) from Wall-E?

Because robots look cooler. Also, I imagine it’s probably doing some stuff with depth perception crap and what have you.

They already do it at POS.

Lots of store websites have a feature where you select a location and it tells you whether a product is in stock or not (and often how much is in stock).

The more interesting question to me is why the robot has to go scan the shelves at the time you ask, when there could just be one (purpose-built) robot that walks the shelves once per night and builds up a model of the entire store that gets used by the other robots for the rest of the day. The hospitality robots wouldn’t even need to be able to see the shelves; they’d just need to be able to guide themselves to the right coordinates in the store and say “it should be right in front of me.”

What hospitality robots are you talking about? Does Walmart have hospitality robots? I only saw the stock scanning robot in the article.

But then you would have to spend lots of money of redesigning ceilings, having regular maintenance work done, etc.


I wonder how many more jobs you would restore by taking away the original retail robot: the cash register. If you just made employees record all sales by hand in a ledger, companies like Walmart would have to hire thousands or millions more people to check out customers. Think of all those beautiful jobs! That would surely make America great again.

You jest, but I actually wonder sometimes if this type of thing would be better for society. The dream of automation and technology was always to "let the machines do the work so we're free to enjoy life and pursue our dreams" but in a capitalist economy it ends up benefiting the owners of the machines primarily, like Walmart and Amazon, not necessarily society at large (though obviously they enable people to buy things cheaper and thus improve quality of life too, at least if people still have some source of income to afford it).

I'd like to think anybody out of a job can just be smart and entrepreneurial and create some other new value people are willing to pay for, but I'm not sure everyone's cut out for that, or how much interest markets have for most of that kind of stuff. Many more small business fail than succeed. Maybe keeping some dumb, honest jobs around would help keep the balance.

the answer of course is "less capitalism" not "more drudgery"

What do you mean by 'less capitalism'?

Instead of introducing more friction into the system so that people have jobs (that are essentially pointless), let's make it so that people don't need to have jobs to survive.

But that will make everyone better off and not just me, why bother. /capitalist

Anyone can own machines. Go buy one. Automate some stuff. Make some money. Stop being a stick in the mud and get with it.

Step one: be rich enough to afford to significantly invest in automation. Step Two: Get Richer.

The problem is that to get that initial investment money you need to work, and the jobs that you're doing to earn that money are disappearing due to automation. So you fall one side of the divide or the other - if you're already rich you'll get richer, if you're not rich your opportunities to become rich are disappearing.

You know a huge portion of the American population is living paycheck to paycheck with no savings right? Most people can't "just" buy a robot and magically join the capitalist class.

Doesn't have to be automated, even. Lots of people start out by owning one snowplow. Or one power washer or one lawn mower.

Try opening a landscaping business. You'll see why this fails.

> The robots are designed to free up store employees' time so they can use it to help customers.

This is probably the most textbook spin I've ever seen. It's as if their PR firm didn't even try.

These robots are obviously being deployed to increase margins.

Personally I would have went for the "adding more value" angle, either by suggesting to the customer that "Your favourite product is more likely to be on shelves", or positioning these robots as "helpers" to shelf-stackers.

> Personally I would have went for the "adding more value" angle, either by suggesting to the customer that "Your favourite product is more likely to be on shelves", or positioning these robots as "helpers" to shelf-stackers.

That's exactly the spin the company that makes the robot is using: http://www.bossanova.com

Depends on whom the message is meant for.

Bossanova's pitch to Walmart: Our robots will increase your margins.

Walmart's internal comms: Our new robots will help our associates save time to focus on other tasks that help our customers.

Walmart's external PR: Our new robot associates will ensure that your favorite product is more likely to be on shelves.

Judging from the pictures they'll still use paper price labels.

Why? It seems extremely arcane to still use paper price labels in 2017, especially for stores as big and centrally run as Walmart.

With digital price labels you can update prices centrally across the whole country within seconds, and it costs basically nothing to implement too, when the labor cost of manually updating paper price labels are factored in.

Imagine if you got your wish (actually, at Kohls it is like this, last time I was there).

Now imagine thru some method, they could tell which shopper was nearby shelves, looking - and who they were. Do a instant credit check lookup (perhaps with something special with the credit bureaus that don't put a ding on the report), to find out what that nearby shopper is "worth".

Then instantly change the prices in that one area of the store - perhaps to gain just a few more pennies from a wealthy customer, or maybe (perhaps) to try to steer a customer to a particular brand who has paid the store for this purpose (by changing the prices of other brands to be higher, so that the buyer might go for the lower priced item if they are on a lower economic rung - or vice-versa to get them to buy a higher-cost item for perceived "luxury").

This would be easily possible with today's technology. It wouldn't be cheap to implement, and likely the back-end systems are so f-d up that it would take a revamp or a lot of work to get everything talking together properly, plus the attendant tech to recognize shoppers (cell phone IMEI? facial rec? something)...

...but I am certain that if the profit motive was there, and it offset those costs, it would be done. And Walmart would likely do it, too.

The offline world would then be on par with the online world when it comes to targeted pricing.

Yeah it's not like Amazon doesn't have behavioral-influenced dynamic pricing. It seems we've passively accepted it.

you can get around most of that using private browsing. but there's no "private browsing mode" in real life.

Isn't it called "burqua browsing mode"? More seriously, of course it's harder in the offline world, I agree.

It is; regular people in the western world - especially men - aren't generally welcome in a shop when they're wearing face-covering masks, for security reasons.

> and it costs basically nothing to implement too.

I find that hard to believe, given the scale of Walmart's operations - they have close to 5000 stores, and who knows how much technical debt they have in their stocking/inventory systems.

Considering the fact that Walmart's last FY earnings were something to the tune of $124.62B,[1] regardless of technical debt or other factors I don't think spending less than 1% of that to do a (probably) much needed upgrade is a hard sell to anyone rational.

[1]: https://amigobulls.com/stocks/WMT/income-statement/annual

Gross profit is not net profit.

You're right, my bad-I updated the comment. Still, even with their net income being $13.64B for the year, it doesn't seem like much of an issue, especially if rolled out over a multi-year period.

I clarified that I meant with labor taken into account.

I don't understand the technical debt argument though - if a robot can scan prices to check that they are correct, a display should be able to display the price without any issue. The prices are obviously accedible in a database.

Shelves need to be checked, faced and re-stocked on a daily basis. Someone is already checking every shelf in the store every day, so the actual labour cost of changing paper price labels is remarkably small.

Electronic shelf labels could facilitate dynamic price changes, but this is likely to be highly unpopular with customers. The mere presence of ESLs could imply volatile or manipulative pricing and undermine customer trust.

> The mere presence of ESLs could imply volatile or manipulative pricing and undermine customer trust.

It definitely does that for me. The heuristic I use in life is very simple - if it can be done and makes economic sense, it will be done. Volatile pricing can be done with electronic signs, and it definitely could seem to make economic sense to some companies.

I'm actually extremely surprised that volatile prices aren't already common in the US. The grocery chains here in Norway update prices across the country as soon as they notice that competitors sell the same products cheaper (provided they want to be the cheapest for that product), often several times a day. ESLs have been very common in big stores since perhaps 2013.

Our BestBuy alternative Power actually had a big ad campaign last year about how their digital signs and their price robot allows them to offer lower prices than what their competitors offer.

>it costs basically nothing to implement too,

Eh. Don't be so sure of this.

Yes, this technology exists. No, it's typically not feasible to implement. The ROI simply doesn't make sense.

Each store houses about 250,000 unique products. That means you'll need that many digital price tags. If they're $4 a peice (highly unlikely), thats 1M per store - and that's only in the new hardware.

You also need to test different environments - will it work in freezers, coolers, and under heated lamps (150F temperatures)? Do those places have wifi access?

Then you need to test and update the wireless infrastructure. Adding 250k devices to a network is not a small task.

Then update whatever software exists to send out price updates. This is honestly the cheapest part, all things considered. And it can be applied to the whole chain / store by store, depends on the roll out strategy.

All in all, it's at least 1 million per store. A brand new store costs the same amount. So you'd also be paying for the opportunity cost of not opening a new store. So you'd also lose out on the potential revenue of that new store.

There's other factors to consider, but the math never shakes out positively. There are bigger fish to fry than price tags.

> Each store houses about 250,000 unique products.


"Supercenters average 187,000 square feet, employ 350 or more associates on average and offer 142,000 different items."

> If they're $4 a peice (highly unlikely)

looks like they're maybe $10 per on alibaba, in lowish volumes.

> You also need to test different environments - will it work in freezers, coolers, and under heated lamps (150F temperatures)?

they can just go check out any of the many places that already use such tags, and note that they work fine.

> Do those places have wifi access?

> Adding 250k devices to a network is not a small task.

they use 433MHz ISM, and don't screw around with IP networking. here's an example device on alibaba


> There's other factors to consider, but the math never shakes out positively. There are bigger fish to fry than price tags.

they're used in hazel's in boulder, co. pretty sure i've seen them used in some other stores, though names aren't coming to mind at the moment.

Would you be so kind as to source some of your numbers? Here it says that a store has 120,000-142,000 products: https://corporate.walmart.com/_news_/news-archive/2005/01/07...

And doing the math here, it appears to be circa $3m in just capex per new store: https://corporate.walmart.com/_news_/news-archive/2014/02/20...

I'm also suspicious of your claim of opportunity cost. Walmart has circa 6k stores and has annual profits in the $15 billion range, and the ability to tap the capital markets as needed. They appear to be able to afford both new stores and price tags.

It would be a huge waste of money because I think the need for electronic price tags will be obsolete in around 5 years max.

I think soon a lot of supermarkets will do same day delivery so you order your food during the day and it gets delivered to you at home the same evening. Or they may implement a system where you order online and you've given a collection time.

Either way going into a supermarket and browsing up and down the isles will become less and less popular so it wouldn't make sense for supermarkets to invest in electronic price tags. I don't know for sure but that's how I see it going.

I suppose because modernising existing systems is tedious, error-prone and expensive. The latter two in principle apply to implementing new systems as well but at least those are exciting and, well, new.

You'd be hard-pressed to find a single company that relies solely on digital processes. In terms of digitalisation everyone seems to be talking about AI, AR, robotics and technologies like Blockchain. However, while certainly interesting those aren't the main avenues for bringing about digital transformation.

Replacing paper-based processes and doing away with the need for physical presence to conduct business are.

So, Walmart will soon have robots running up and down their aisles while at the same time we still largely rely on paper for invoicing and accounting.

We'll soon have self-driving cars for all our wonderful commuting but we still haven't figured out how to use technology for remote working to make travelling to a ridiculously expensive place called 'office' a thing of the past.

Probably because it's the lowest common denominator across all states to follow codes.


A company the size of Walmart would need to employ many lawyers just to deal with state to state pricing laws.

Paper doesn't break. The price on paper doesn't suddenly disappear because the internet is down or someone's program glitched, which is especially important for unit priced items which have strict laws about accuracy in pricing. You don't need to bug the home office everytime you need to change a price locally or they make an error in pricing because someone transposed a zip code number.

I mean...we have something like this already with climate control and thermostats in offices and stores. The central control can screw up spectacularly at times,

At the very least you will need another little robot or human that goes by and verifies that the prices have actually been updated. I can't imagine the LCD price displayers will be super high reliable and built to last decades.

You seem to think that electronic price labels is a new unproven technology, while it is actually extremely common and well tested. After the initial introduction period, where errors can occur, it's pretty much guaranteed to work as flawless as it does everywhere else.


You guys think there will be a backlash to this? Some sort of modern luddite movement? Whenever I shop I refuse to use the self-checkout, you think that sort of thing might spread?

Why do you refuse to use the self checkout?

I prefer self checkout in theory, but often the machine doesn't work properly, and you need a human to login and fix it anyway.

The whole thing where it checks if something is in bagging area seems to be pretty buggy in most of the stores where I've used them.

I don't use self checkout because I'm lazy, tbh. It's more work I need to do. Particularly at the grocery store I go to, I know which cashiers are good and efficient. If I have a coupon, or if I'm buying a clearance item, it's also not a big deal, but self checkout always seems to require a "help is on the way!"

Convenience/efficiency more than anything. Most of the time, human checkout is faster than self-checkout. I can prepare my payment in parallel with the human operating the checkout hardware. I don't have to learn and navigate the touch screen menu system (even if it can be learned in a second, that's one second more than the zero I get with human checkout). Often, self checkout kiosks fail in ways that require a human to be summoned anyway.

Prepare your payment? What's there to prepare?

Pull out wallet, extract credit card, swipe or insert chip, wait, sign, put back card, put back wallet. All these things can be done in parallel while the cashier is also working. With self-checkout unless you are very good you can't do this while you are also scanning all your items.

Ah, right, I forgot about that. Here you can't "swipe or insert chip, wait, sign, put back card, put back wallet" before everything is scanned, since you validate the total before inserting the PIN, so we can't save much time by avoiding self-checkout.

Can't answer for GP, but I use human checkout to provide epsilon demand for cashiers who need the job. Efficiency isn't everything.

I used to feel that way about bank tellers (having been a former bank teller a long time ago...) then I realized I was wasting 20 minutes in line for a simple deposit that the machine could do in under two minutes without a queue.

True; a 20-minute difference would definitely be enough to tip the scales in favor of self-service. But in the grocery stores in our area, there are almost always plenty of registers open and little waiting. So the time requirements being approximately equal, I'd rather pay a skilled cashier and bagger to do the work of speedily "ringing up" and bagging my groceries, while I stand there and chat with them. (As a teenager I had a summer job bagging groceries one year, so I can relate.)

I'm still baffled that baggers exist. From my (German) perspective, the entire concept of employing a human just for this task is absurd.

The 20 minute wait was excused by your willingness to accept the offer of automation.

I visited a Walmart last week and notice they had converted 1/3 of the normal checkout aisles to self-checkout stations. They left the conveyor belts in and I noticed how people still used the conveyor belts the same way whether there was a line or not, unloading each item onto it before scanning them all.

Don't you feel a tinge of guilt when you walk in one day and now there's only two tellers instead of three? Maybe I'm just weird but I was at target yesterday and there were only three aisles open with lots of people going to self checkout. I said fuck it and waited in line for a person.

Do you go down this rabbit hole?

How do you pick between buying wheat and taking it to the miller (preserving his job) and buying flour from the grocer?

Do you seek out gas stations that still pump gas for you?

Do you worry that buying a fuel efficient vehicle or insulated building can have a negative impact on people working for energy producers?

>Do you go down this rabbit hole?

What requires me to go all the way down the rabbit hole? Everybody has moral beliefs that they don't follow to their extremist conclusions. I see people getting laid off or turnover being allowed to happen - I see that the replacement job for a cashier gig is sometimes even worse - I do my little part to make society a little less bad. I also don't feel that pure individualism is really the way to live so I do sacrifice a bit of a wait in line and hopefully it helps the girl that checked me out keep that gig for a little bit longer.

Nobody wants to be a cashier for 8 hours a day though. The only reason we see losing your job to automation as a negative is because our welfare and education systems are broken.

All you're really doing is creating more pointless work to treat the symptoms.

>All you're really doing is creating more pointless work to treat the symptoms.

If pointless work is the best we've got that's fine by me just as long as someone's getting paid on the other end.

Besides if you're sick and know you won't be cured for quite a while you might treat the symptoms as well. Treating the symptoms isn't bad as long as we don't forget to also treat the cause.

My comment is sassy, but I really was curious about the answer to the question.

My answer is self-evident. I go partway down the rabbit hole. However deep "refusing to use the self-checkout" is.

Sure, I thought you answered it in your other comment, I was just explaining that I hadn't posted just for the sake of being antagonistic or whatever.

Oh I see. Yeah my initial response was a bit rude, sorry about that.

I refuse to feel guilty, but I'm happy to blame the large soulless corporation that created this moral dilemma for me in the first place. I didn't fire those people. I don't eat at McDonald's anymore because I want to stay alive for a while, but I'm not going to feel guilty about their workers either. I'll admit standing in line for my robot checkout while idle cashiers stare at me with forlorn sad eyes makes me uncomfortable. And yet I have no issue taking money from an ATM even though there's a building full of money-counting humans behind it.

> I'll admit standing in line for my robot checkout while idle cashiers stare at me with forlorn sad eyes makes me uncomfortable

What? I just assumed that everyone optimized for the minimization of time spent waiting in line. Personally, I'd never wait in line for self checkout if a cashier was idle!

No I don't feel any guilt in that situation.

If there was a 1% service fee for using the cashier (or a 1% rebate for self-checkout), would you still use a cashier?

> If there was a 1% service fee for using the cashier (or a 1% rebate for self-checkout), would you still use a cashier?

Definitely. Our average grocery bill ranges from maybe USD $50 to $125 (we're empty nesters), so 1% of that would be noise level. Heck, back when baggers accepted tips, I always tipped at least $1 (when I was a teenaged grocery bagger it was for tips only); that's more than the 1% would be.


I like your use of "epsilon." I have a lot of little rules I like to follow, like the Nestle boycott, that are just drops in the bucket. But it's nice to have a set of personal standards to follow even if they are inconsequential at scale.

I'll be using epsilon to describe these quirks from now on :-)

I DID the cashier job at the store I go to, and am still fairly familiar with the software/hardware. If I could just step up to a normal register and run through my stuff, it would take seconds for me to check out, as it does to check out small orders when you have a cashier who somehow manages to care at $10/hour.

Meanwhile, self checkouts are slow as hell, cumbersome, and double check everything you do, but still manage to not stop theft. Not to mention many purchases need a human in the loop anyway, and the amount of times other shoppers spend ten full minutes also fighting with the things, but without the understanding of the back end POS software, and you could easily replace a group of 4 self checkouts with the one person being paid to watch them anyway

People need those jobs so I want them to exist.

The 18th century version of this was not using an umbrella to preserve demand for coach drivers:

"Over the years, Hanway and his umbrella fell victim to all sorts of abuse from Brits he passed on the sidewalk. The most pernicious abuse came from an unlikely source: coach drivers. In England at the time, hansom cabs (two-wheeled, horse-drawn carriages) and sedan chairs were the primary modes of transportation. Business boomed especially on rainy days, as both hansom cabs and sedan chairs came equipped with small canopies that kept passengers dry. When it rained, Londoners flocked to these coaches, so Hanway’s umbrella represented a threat to business.

Fearing an interruption in their personal incomes, many hansom cab drivers and sedan chair carriers grew violent toward Hanway. According to the British history magazine Look and Learn, when they saw him walking by, they often “pelted him with rubbish.” On one occasion, a hansom cab driver even tried to run Hanway over with his coach. Hanway reacted by using his umbrella to “give the man a good thrashing.”" - https://www.atlasobscura.com/articles/the-public-shaming-of-...

And if I lived back then I might have taken the coach instead of using an umbrella. I don't understand your point.

Because I drink a lot.

Can’t buy booze in self checkout (in CA).

I'm in Austin, TX and I can buy booze in self-checkout, as long as it isn't Sunday morning.

Do they scan your ID? If not, I'm surprised someone hasn't come knocking about this...

Sadly, it's not that simple. Facial verification has come a long way, but it's not accurate enough. You need a person to verify the person's ID matches the face of the customer that is buying it - otherwise some teenager can swipe his Dad's ID and buy alcohol.

Plus you need someone to verify the customer is not already drunk - selling to someone who is already visibily intoxicated is a great way to a lawsuit.

It's obviously not the same thing as a self-checkout machine, but there are automated border control systems in some countries (e.g. Germany). Scan your biometric passport, step into the mantrap, look into the camera, welcome to Germany! If it's good enough to let people pass border control, it should be good enough to sell alcohol. The technology exists, I just wouldn't want to use it at a grocery store.

In washington you have to wait, a long time, for a clerk to come over and check your id and unlock the checkout machine. it's obnoxious

No. There's one attendant for the group of machines and as soon as the 'enter birthdate' pompt appears on the screen, it gets cleared. I think the attendant is just watching and I'm clearly over 21. It usually is cleared so quickly I don't see it.

I'm not the one you're asking. But I work with computers all day. Even for an introvert, human contact is nice every once in a while. So I use the human cashier. S/he may be less efficient than self checkout, but the self checkout machine never smiles at me...

It kind of saddens me that people have to point to human cashiers as a source of social interaction. Maybe you can find a place where you can be with people who accept that you want to be solitary at times. I attend my local hackerspace for that purpose.

I'm going to guess because of how symbolic it is regarding our society's relation to labor.

I refuse to use it because they don't provide a discount for doing so. What's my incentive?

For one, Wal*Mart doesn't provide cash-back at self checkout

Yea it does...and people often forget to pick it up

I love bagging my own groceries the way I like, and avoiding interaction with a potentially disgruntled and inattentive cashier. I love self-checkouts.

oh? you don't want your bread to be a cushion for you bananas which are there to keep your canned goods from falling through the thin paper?

I think you're going to see a lot of kids knocking them over for fun, at least at first.

Given the previous reactions to any sort of move by fast food chains toward anything like automation, I can't imagine this won't be picked up as fodder for the "look what happens when you ask for a living wage" contingent.

You don't think wages have any effect on the decision to automate?

I will always vote for raising the minimum wage but at the same time I know it probably makes automation more attractive.

I don't think it has as extreme an impact as the people posting images of a self-service menu imply it does. I will concede that a lot of my aversion to the argument is from the disdain for people with lower-class jobs that I read into it.

But while I agree there is an impact, I also don't think we should avoid automation simply to have someone doing a mindless job if it could effectively be automated. People don't necessarily need jobs, they need the resources that having a job allows them access to.

I think we'll just see a ton of videos of electric scooters crashing into them...

If by "backlash" you mean 'looks like late 18th century France,' then yes, I think there will be a backlash.

> The robots are designed to free up store employees' time so they can use it to help customers.

If you’ve ever been to a Walmart at 2am, you know that this isn’t true. A lot of people will lose jobs over this.

I wonder if these robots will suffer from the "roomba dog mess" problem. Meaning if a jar of something gets dropped in the aisle, will the robot happily drive through it and then leave tracks of it throughout the store?

I'd also be interested in learning about the false positives. For example a product get turned around on the shelf or returned askew does that trigger an alert? And is that a good or bad thing?

I guess the proof will be in the pudding. If these are still in use two or three years from now then we know they were a success.

From a personal perspective I'd love these to improve me locate products within a store. Many stores still fail to tell you which shelf a product is on on their mobile app, and the few that day don't give you a map of the store showing you the literal location of it. I'd love to have a "Google Maps"-like experience with locating an item.

It seems interesting to me, that computers seem to be taking on managerial roles and humans are left with manual labor. The robot knows how to tell a human to fix a problem but doesn't know how to restock the shelves itself. Similarly, in warehouses, computers tell humans what to do, and the human's job is basically just to grab things.

In the end, the robot is already higher on the ladder than the humans.

Its not that we cannot build robots that restock shelves, it is that shelves are designed for humans to stock. And emulating the range of movement and dexterity of human hands and the sensory data we get when interacting with such shelves is really expensive, rather than prohibitive. Especially to maintain. Think of all the fine motor joints you would need in such a robot that could break all the time (like how fragile human fingers are). But how else are they supposed to stock individual bottles of maple syrup with weird handles while stocking jars or bags of flour a shelf over?

In situations designed for robots, like Amazon warehouses, they don't have a problem stocking shelves as long as the range of required behavior is much more restricted so the instrumentation can be more rugged and less articulate.

The newer Amazon warehouses actually provide a nice example of some further refactoring to make at least parts of the stocking and picking more automate-able, in that the shelves all move and he humans stay stationary. Fingers are still hard, but moving across a warehouse is easy, so may as well shave off that piece of the task.

It'd probably be better to design standard packaging and automated shelving systems that could be easier for robots to stock. Even better? Make the shelves robots themselves, so they could re-arrange themselves as needed, and go and "fill up" for stocking in the back, etc...

Reminds me of this story: http://marshallbrain.com/manna1.htm

I always thought it was amusing when people said it would replace burger flippers and the like. Why? They're cheap, and that's relatively expensive/difficult to automate. The slightly higher paid manager though? Ripe for automation.

"The girls liked it because Manna didn't hit on them either."

Interesting story. I'll be sure to read it.

> They're cheap, and that's relatively expensive/difficult to automate.

that's a weird thing to say, because burger flippers ARE automated -- 99% of the work of making a McDonald's burger is already automated, only the remaining hard-to-automate 1% is left for the humans.

Same here....I suspect it's middle and lower managers that will be first to be obliterated.

Cashiers don't simple run the register, they do other stuff. Janitors fix stuff and carry heavy boxes, etc.

Managers and supervisors? Overpaid if they're doing the above or doing something can be automated.

Reminds of me of the Warren Bennis quote: The factory of the future will have only two employees, a man and a dog. The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment.

>The robot knows how to tell a human to fix a problem but doesn't know how to restock the shelves itself. [...] computers tell humans what to do, and the human's job is basically just to grab things.

Yep. There's a name for that observation: https://en.wikipedia.org/wiki/Moravec%27s_paradox

I think this could be said for many technologies: the cashier scans barcodes, the cook at McDonalds flips burgers when the alarm goes off, the anesthesiologist or pilot watches gauges...

And Uber tells the drivers where to go.

Near my house, 4 red octagons with a simple white drawing command humans to stop and cooperate.

I don't understand what you mean.

Stop sign. I don’t think it’s a good analogy.

Actually, I think that anarchists would tell you it is a great analogy. They would claim that we are being controlled by a system of laws, a system which is like a computer program written by novice programmers who do not know how to debug, and who paid no attention to the finer details. That not only is the legal system like a robot which controls us, but it is like a stupid, primitive, and brutal robot which controls us.

Specifically a four way stop, and I'm not sure I agree that it isn't a good analogy. It's an autonomous system (traffic laws) commanding human behavior.

In South Africa, these are literally called robots.

edit: was too fast, sorry. Traffic lights are called robots in SA.

Managers make decisions when the information is equivocal.

That is very hard to replace.

Resource allocation and unequivocal decisions are for the machines to help us with.

If, as a manager, you do more of the latter: be afraid

> If, as a manager, you do more of the latter: be afraid

As a gardener, I have never feared a robot that could pull weeds. As a home-maker I have never feared a robot that could wash dishes. As an inventor, I have never feared a robot that could draft diagrams. And yet those people who are at the mercy of the market fear, and this is why, I believe that capitalism is headed for either disaster or elimination. Capitalism has created a problem which does not exist.

Post-scarcity is only a problem in capitalism.

I'll take it. Because "how to fairly distribute all of this stuff that nobody had to work to make(1)" is a better problem than the one it replaced: "how to choose who should die because we don't have enough".

(1) Just so long as the answer isn't "make them die because we don't need them anymore".

... because only capitalism has made post-scarcity possible.

The monarchies of the late 19th century were technologically advanced, but that did not stop Europe from adopting democracy in the 20th century.

The monastic system was the main scientific institution for over a milenia, both in Europe and China. Only an ignorant person argue that this system developed no new technologies.

You say "only capitalism", but what are you comparing capitalism to? In the past 200 years, capitalism has been the only economic institution in the world, how do you know that other institutions would have been incapable of technological advancement?

If your point is to compare capitalism to state socialism, then you would be far off the mark. The USSR was extremely technologically advanced. The hunger was due to other problems, mostly having to do with Stalin intentionally creating famines in order to weaken insurgency.

"Is has been proposed that the Soviet leadership used the man-made famine to attack Ukrainian nationalism, and thus the man-made famine may fall under the legal definition of genocide.[57][58][54][59][28][60] " https://en.wikipedia.org/wiki/Holodomor#Causes

I am not promoting state socialism though. I would like a future in which each person/family unit/group of friends owned a small amount of land, and robots that would help them. Quite the opposite of state socialism. In state socialism, wealth is concentrated in the hands of the government. State socialism is basically monopolistic capitalism, and I believe that the two systems share many problems.

I want the distribution of wealth, not its concentration.

Communism works when the communist dictators don't commit genocide or run a brutal police state.

Its such a good system it needs to be implemented and maintained by force.

I didn't claim that "communism works" I said that it was technologically advanced and capable of attaining post-scarcity. As a counter to castle-bravo's claim that only capitalism is capable of developing the tech required to attain post-scarcity.

Ah, there you got me wrong. I said that only capitalism has made post-scarcity even remotely possible. We're seriously discussing the possibility and implications of full automation within the next fifteen years; that never happened in the Soviet Union, their government collapsed first. In principle, any society with enough scientists and engineers could implement full automation.

As a side note: post-scarcity is really only possible in virtual worlds like MineCraft. There's a limited amount of land, people's time is limited, people's attention is limited, and what other people will participate in is constrained by their desires and their ambitions. Unless somehow compelled, there's nobody who will set me up with a nice automated farm where I can live out my days playing mandolin and reading mathematics as I would like when they could set themselves up and not have to hear my racket across the fence. Though it's counter to our desires, we are forced to compete for resources ... even if most of us aren't struggling for survival.

> In the past 200 years, capitalism has been the only economic institution in the world

This is not true.

Ok, I'll admit that some remote tribes were not. But even the USSR was a market economy.

if market is your model of choice, sure, and if your model of russia is simple enough, too.

Still, if you eliminate the easy stuff, you might get away with one manager for a whole region.

It's not too different from a static type system. Humans are great at general problem solving, but computers are better at precision (but only when you can precise measure something).

Counting inventory and generating pick lists isn't really "managerial". The word usually means managing people, which still requires human judgement.

More like: humans who program the robots > the robots > humans who are instructed by the robots.


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact