Stop and Shop now has big, goofy-looking robots patrolling its aisles | Hacker News Reader

Stop and Shop now has big, goofy-looking robots patrolling its aisles

newfoodeconomy.org

34 pointsscott_s86 comments

I was in BevMo a few weeks back and there was a similar robot. It had a screen on the front to look up what you’re searching for and it would bring you to it.

I decided to give it a try since I was trying to find watermelon flavored liquor for the SO. I found some pre-mixed watermelon margarita on the screen and then it proceeded to bring me to the right spot....or at least it tried.

On the way to the aisle and shelf, it clipped the corner of a chest-height shelf rack knocking over a ton of whiskey on the top shelf shattering on the ground.

It was seemingly unphased by hitting the shelf, but right after ran straight into a ceiling height rack of shelves knocking over bottles of tequila.

It just stayed there with its face buried in the shelves and simply said, “Can I help you find anything else?”, despite not being where the pre-mixed margaritas were.

There was a single employee there working the register. When the whiskey crashed, everyone in the store just watched and then when it crashed into the tequila, the employee looked at the mess from behind the register, sighed, and then just continued ringing people up. This clearly wasn’t the first time she’s witnessed this thing make absolute chaos.

I’ve yet to go back to BevMo since then, but I imagine they’ve gotten rid of it.

We have one in our local Stop and Shop. It has a creepy smiley face plastered on it. It's easy to say that it's super weird and suspicious, gets in your way, likely not cost-effective (at least for its ostensible purposes, who knows what slimey face tracking or whatever it might be doing), which is all true.

But on a slightly positive note, my kids 5 and 7 love it. They pretend like it's a killer robot chasing us, which it oddly does seem to do. No matter where we go in the store it's always popping up around the corner. They play hide and seek with it, try to surprise it, hunt it down. Good training for Judgement Day. ;)

2020's turing test: a robot/drone is intelligent when you can no longer distinguish it from a murderous stalker / killer clown.

Does it have an e-stop? Can you hit the e-stop?

Not sure what you mean. Emergency Stop? I know it's bumped/tripped my kids on a few occasions, but they were asking for it. :)

It should have a big e-stop on the robot. This is to insure that if it runs over kids or malfunctions then you can shut it down.

If you work for a public company, you have the ability to get the Real Deal on their plans.

From TFA:

> Although Jennifer Brogan, Stop & Shop’s director of external communications, assured New Food Economy that the robot is not meant to replace workers, Ahold Delieze has explicitly told shareholders that the company is investing in automation and artificial intelligence to supplement or even replace human labor.

Management lies to employees all the time about future plans, staffing levels, job roles, etc. But they can't lie to their investors. (Or, they usually share the worker-hostile plans more easily)

So if you're curious about your company, dial into the earnings calls, read the investor relations press releases, etc. That's where you'll get the better info.

I don’t think one needs to do all of that to understand that people prefer to spend less money than more money for the same thing. If you’re not expecting your employer to be looking into ways to reduce costs, just like everyone tries to reduce costs in their personal life, then I would say you’re intentionally oblivious to reality.

Depends. A business can be focused on creating value for its shareholders and owners, or it can be focused on creating value for its employees and owners.

If you don't view the purpose of a business as needing to extract and funnel wealth upwards it doesn't need to come down to spending more vs spending less, spending more might even further the goal of the business if you view its purpose as creating jobs for people.

Essentially, it is possible to consider employees as an asset instead of a fungible commodity.

> A business can be focused on creating value for its shareholders and owners,

What is the difference between shareholder and owner here?

In a low margin commodity business, such as retail and grocery stores (that don’t cater to top 10% by wealth), it absolutely does come down to spending less.

Proof is the success of Walmart, Amazon, Aldi, Costco, Dollar General etc. Customers don’t care about anything but buying goods at the lowest price.

This is a great example of how cost cutting doesn't always mean rock bottom wages. Costco and Sam's Club are good comps. Yet they're way different when it comes to their workforce strategy:

http://www.slate.com/articles/business/moneybox/2008/06/wage...

I don't think it's naive to wonder about future staffing plans or employee relations. Some businesses genuinely see an alignment between profitability and hiring, or between profitability and increasing their employee morale. Others will cynically parrot these things as "Important" but will not really treat them as such.

To find out what category your company belongs to, it helps to see how the conversations they have with owners differs from those that they have with the workforce.

I once worked for a company that was honest to a fault with all of us. Potential layoffs were telegraphed months in advance. The things they told us they cared about happened to match up with their subsequent actions. I've also worked for places where the internal employee communication was so clearly disingenuous that we started to just invert whatever was said. That had a material negative effect on the business, as all the best talent left when they could, leaving an organization that just couldn't perform (and profit!) as well as it could have.

All it does is monitor for hazards and report them so that a human can address the issue. Nobody is going to have a practical and affordable robot that can restock shelves any time soon.

> So if you're curious about your company, dial into the earnings calls, read the investor relations press releases, etc. That's where you'll get the better info.

Good point. Lying to your employees or customers is unethical, but lying to your investors is illegal (securities fraud, and "everything is securities fraud", as Matt Levine frequently points out in his Bloomberg column "Money Stuff").

https://duckduckgo.com/?q=money+stuff+levine+%22everything+i...

The long game is probably a litany of other vision-based tasks including things like stock management. Not knowing what is even on their shelves is a huge issue for retail— having oversupply is wasteful and can result in spoilage, and undersupply means lost business and unhappy customers.

Note that there are other companies getting into this space:

http://pal-robotics.com/en/products/stockbot/

https://shelfierobot.com/

In fact, that's exactly the purpose of the one in our local grocery chain's stores (Schnuck's in St. Louis). It scans the shelves for inventory.

It's not as big as the one shown in the article and I don't see it all day, every day, though I wonder if I've just gotten used to seeing it and now am blind to it.

https://www.youtube.com/watch?v=qZBeV4hHAak

Thanks, Tally was the one I was missing. Too late to edit the post, but yeah: https://www.simberobotics.com/

"We are partnering on technologies that are not only helping us make the customer experience even more personal and relevant but also operate more efficiently and manage labor shortages in our markets"

That's laughable. I wonder what's the top management's perception of working on the floor. From the article it seems that this robot is practically useless (glorified Roomba that doesn't vacuum) and am sure the management has a different plan in mind.

I walked into my local bank branch last week and couldn't find a withdrawal slip. One of my vendors only accepts cash or check. I haven't used paper checks in years and the ATM only dispenses $20's: I hate carrying a thick stack of bills.

All the slips had been removed from their usual place so I walked up to a teller to explain the situation. He said, "oh, we implemented a new paperless system." And then pulled out a withdrawal slip and filled it out for me.

Me: "soooo, your new paperless system is paperless only for me I guess?"

Him: "yeah, I guess they figured we needed something to keep us busy."

So yeah, I can totally believe that management's perception is completely disconnected from reality.

I just use a local credit union but it hasn't occurred to me to even bother filling out a withdrawal slip for a long long time.

I forgot this was even a thing. Whenever I need large bills to buy something on Craigslist, I just walk up to the teller, give her my ID and Debit card, and ask for what I want. And that's just a normal bank, not a credit union.

It seems pretty likely they are going to use it to keep track of stock (the article notes that the manufacturer intends the robot to have this capability).

While they're monitoring spills for now, these robots are also designed for inventory. That's why they're as tall as the aisles.

I'm sure the robots are designed for something better than monitoring spills but they're only used for that. And most likely they won't be monitoring inventory any time soon.

Also, I don't understand why monitoring inventory has to be done this way, they already know what new inventory comes into the store and what goes out when the cashiers scan the items at checkout. Seems to me like the simplest solution is to subtract the two (actually a tiny bit more complicated than that since there's a running inventory). Just querying a database you can see what's on the shelves, what's expired, etc by tracking all the items virtually. And we're not talking about monitoring an amazon sized warehouse, it's just a supermarket..

This doesn't tell you what your shrink is, though.

"We made your experience more personal by eliminating people."

It's mentioned in the article

Direct link to the company behind it:

https://www.badger-technologies.com/

It seems like the primary focus of this device is inventory + planogram management. That it can also scan for spills + messes seems like an afterthought.

You know what would be better? 2 more checkout aisles open. The ridiculous amount of time it takes to check out is the #1 reason I avoid that chain. It's not unusual to have to wait 10+ minutes to checkout because only 1 or 2 of the 15 checkout lanes are open.

But that would mean having to hire more checkers, and no ambitious mid-level executive is going to get a promotion from doing something as mundane as hiring more checkers. (It may actually be the opposite of a career enhancer. "Hey, our expenses went up .37% last month! Find me the person who's responsible so I can fire him!")

"Deployed and managed a fleet of autonomous robots", on the other hand, looks much sexier on the résumé. That executive is on the fast track to the C-suite.

OpEx vs CapEx

I can’t for the life of me understand how vision related tasks should be done by roving versus fixed cameras.

Even if you need higher resolution than you can get from a wide field of view, a fixed camera that can pan and zoom then?

The only thing I can think of is there’s zero upfront installation cost, no need for a crew to wire anything. Ship the robot, plug in the charging station, and it’s off?

Maybe the ancillary reduction in shoplifting, even if it’s not actually designed to monitor that, pays for the robot.

Here, it’s probably the shelves and isles. Cameras mounted high have trouble down low, front can’t see the back, etc.

They probably don’t need down to the second data, and a dorky robot gets people talking about an otherwise forgettable store.

They have the backend hooked into the PA system for announcements when it finds a spill.

Dense shelving can have a lot of blindspots, that's one of the reason those Amazon stores are absolutely plastered in ceiling cameras; that being said 700 CCTV cameras would probably still end up cheaper than the proprietary robotics system it seems they are using now

Well for one they're easily replaceable, infinitely configurable, and don't necessarily require someone to watch them

I have one near me. It's intrusive & annoying, often blocking aisles and getting in a customer's way in a more dangerous fashion than the situations it's supposed to detect.

I also have one near me. I don't find it intrusive or annoying, and while it may get in a customer's way now and then I do not consider it dangerous in the least. I often find it sitting in front of a spill of some sort in produce trying to page a worker to come clean whatever it is.

Are spills so common in grocery stores that this is really worthwhile? Sure, I've seen spills, and staff cleaning them up; I've even reported a spill to staff when it didn't seem to have been noticed yet. But it's hardly an everyday occurrence, in my experience.

Not sure where you shop, but I can't remember the last time I went to a "normal" (i.e. not Whole Foods level overpriced) grocery store near me and didn't see some kind of spill, mess, or hazard on the floor.

Granted I specifically avoid shopping on weekends but I go to two grocery stores of a chain known for being value priced in a city known for being the opposite of upscale. I can't remember the last time I saw a spill or something on the floor that was actually worth avoiding, a green bean someone dropped, sure but nothing worthy of being called a "spill".

It sounds like you've never worked in a grocery store. I did for 6 years (all of high school and 2 summers during college.) When we weren't on the register, we were almost constantly mopping from spills and broken jars. In a store with older equipment, even the coolers in the produce and frozen department will cause minor condensation on the floors (most grocery stores have rubber mats in those areas specifically because it would be a full-time job keeping up with it otherwise.)

It was certainly a t least a 2-3 times a day occurrence.

The reason is pretty clear: grocery stores work on such slim margins that even a minor lawsuit from someone slipping or being cut on broken glass might be the end of the store.

When this robot was added to my hometown's grocery store my family was similarly mystified discussing it, but my father is an insurance professional and said it's a really serious liability expense for grocery stores to manage spills. People who slip and fall on legitimate accidental spills that weren't attended to promptly will sue the grocery, and there are people who will create spills to self inflict injuries for the same reason. His opinion was this kind of oversight is a no brainer for a grocery to reduce the liability overhead, but we also couldn't determine why that's better accomplished with a robot than cameras in the ceiling. Maybe to reduce perceived "big brother" vibes?

Edit: I don't think your comment should be downvoted below the fold, because I think it's pretty natural for people unfamiliar with the operation of a grocery store to have a knee-jerk skeptical reaction to putting a robot in one

I can see how this could, theoretically, help manage spills and related injuries. However, at least the one near me is so slow. It would take hours for it to do a round of the store. It spends most it's time standing still in an attempt to not get in the way of passing shoppers. I'd think that having a stock clerk make a round of the store every 45 or hour would work much better. A brisk walk across the top of the aisles, looking down them for spills, wouldn't take more than a few minutes.

Spills happen all the time, and if they are not cleaned up promptly they create a liability for the store. People (old people especially) can be seriously injured or even killed by slipping on spills. That could cost a store millions of dollars in injury or wrongful death damages per incident.

I've seen the one near me make turns into people's carts, not stopping until something crosses it's front sensor. Or stop in the middle of aisles preventing people from getting by on either side. These issues seem more prevalent when the store is busy.

I work for a company that's in a partnership with Ahold Delhaize. While we were doing some in-store tests of our software product with them over a year ago, the topic of the robots came up. As they say in the article, the main purpose really is just to scan for spills and obstructions. That being said, I believe the primary driver is insurance-related. The stores are required (or perhaps given a discount as incentive) by liability insurance providers to regularly check the aisles for fall hazards, and the required sweep frequency is pretty high. In a big enough store you'd probably need to dedicate someone to it full-time--I guess in this case they opted for a googly-eyed robot.

It's like a giant "Clippy" following you around and doing nothing other than being annoying and intrusive.

I don't need these in my life!

Am I missing the point somehow? Why is this not implemented as a series of super high def camera disguised as surveillance cctvs? They could spot spills just as effectively, no?

> “It’s really not doing much of anything besides getting in the way,” said Kristen. She complained that the robot pages the store nearly constantly, over the smallest things—like a stem from a bunch of grapes—and it will go around in circles until an employee comes to clean the “hazard” up. “Right now he’s a glorified Roomba and he doesn’t even vacuum.”

Market opportunity spotted - are there industrial Roombas for retail spaces? Even better if they can do outdoor sidewalks.

> Market opportunity spotted - are there industrial Roombas for retail spaces? Even better if they can do outdoor sidewalks.

It would have to work much better than a Roomba. Random cleaning won't cut it, it would have to be done in a methodical fashion.

Could such robots be built, and be made safe and effective. Probably. But the price point is going to probably kill it.

Think about it - this thing costs $35K USD - now you want to make it sweep things up and do it in practical manner, etc - that price is going to rise significantly.

But - if you can do it, and keep the price low - then you might have a chance.

It's definitely a space I'd love to explore, but I'm not a businessperson or a real entrepreneur (I definitely probably have all the parts needed to build such a robot in my shop, though - including more than a few old Roombas to boot).

Marty sounds like a big waste of money to me. At the very least they should have made it able to sweep or vacuum up what it finds on the floor.

Slip & fall claims are a big liability for stores like Stop & Shop. One avoided per year pays for Marty.

Marty seems to be unreliable and produce a lot of false alarms. At some point crying wolf won't be taken seriously anymore and Marty's warnings will be ignored.

An alternative to this is to hire a person to babysit Marty, follow him around the store and whenever there's a real spill to clean it up for him. /s

Wouldn't a few HD cameras on the ceiling and OpenCV be as effective at 10x less cost without getting in everyone's way?

That's what I was thinking. Seems like they will want to expand the capabilities, though, and that may be hard to do from ceiling cameras.

Unless its going to physically stock the shelves, motion is so expensive and difficult to maintain and cameras and other sensors are so cheap, I can't ever see a mobile robot being more cost effective than a sensor blanket.

It makes natural sense for the robot to move from identifying spills to cleaning them up. Lots of people say "well somebody just sees a spill and sends a human to clean them up" which doesn't work when the first person to detect the spill is the old lady who slips in it and breaks her hip.

If Marty is roaming the aisles at a reasonable speed (i.e. not so fast as to endanger shoppers!), there's still a substantial chance that the first "person" to detect any given spill is the elderly person with the fragile hip, while Marty ambles around the other side of the store.

So how many robots are needed per store if they're to have more effective coverage than asking the employees who roam the aisles tidying and re-stocking shelves, helping customers find things, fulfilling click-and-collect orders, etc., to also report spills?

Or perhaps these stores don't have any such employees on the floor -- they're just an array of unserviced aisles, with some staff at the checkouts?

>If Marty is roaming the aisles at a reasonable speed (i.e. not so fast as to endanger shoppers!), there's still a substantial chance

Still substantial but less than without it.

Most of the restocking and cleaning happens when the store is closed. Customers don't want to have people getting in their way stocking shelves and you don't want to be mopping floors (other than cleaning up spills) when customers are walking down them.

Of course they'll want to "expand the capabilities". With facial recognition, and their database of you and your family's shopping habits, it'll be happy to sidle up to you with helpful suggestions: "Your wife's phone says she's driving like she's really stressed this evening; why don't you pick up that bunch of flowers to help her feel better? And she'd love this wine with dinner. Wait, that's not the shampoo she usually buys; get her the one with coconut oil."

If they wanted to use it to check stock levels/facing then cameras above wouldn't be quite able to see that.

Cameras with wide angle lenses every 10 feet on opposing shelves. Still cheaper. Much less maintenance. Software assembles a seamless view of every shelf in the store in real time. No need to wait for the bot to wander by.

Now you've created the need to re-wire shelves for power + data, map the cameras to store positions, keep people from stealing them, maintain/replace them, etc. I'm not at all convinced that would be cheaper in fully burdened cost than a few roving robots.

Cameras every 10 feet on each side of an aisle. Now you're talking dozens and dozens of cameras. Definitely not less expensive than the glorified Roomba they have roaming the aisles that I've seen.

>Another said the robot is currently broken and stored on the charging dock, a $35,000 decoration.

Thats a lot of cameras you could buy right there. I was figuring around 100 for a medium sized store. (40 ft isles, 12 of them, 2 cameras every 10 feet). How much is a little wifi camera that can send back a few images per minute?

>How much is a little wifi camera that can send back a few images per minute?

I was discussing cameras with someone who derives a non-negligible portion of their income from streaming 4k video. The "gold standard" as far as they are concerned is a $60 Logitech webcam. Swap out the cord for a wifi chip, swap out the consumer housing for something that clips to a shelf and can be secured with a self tapping screw and power it with a cheap solar panel and I still can't see it coming in above $100.

In quantities of a few million. If you discount any development costs.

More than the robot. You're not going to pay the price of $100 wifi cameras on Amazon, just like you're not paying $35,000 for a glorified Roomba.

Walmart already does this in certain isles to deter theft

How would these cameras identify spills? I would imagine it would be hard to detect a clear liquid spill from a ceiling camera, but maybe I'm wrong. I would imagine if ceiling cameras could do this job, they would, as it seems way less expensive and intrusive.

Can the bot reliably identify clear liquid spills?

Even if it can't, it has wheels and, I assume, distance measuring sensors/encoders, etc. Slip is "easily" detected by monitoring wheel response to motion commands.

Polarization imaging [1] could probably do it. Granted, the sensors are expensive and lower resolution than one might like.

Another way is to measure surface specularity. Use one or more IR spotlights at different angles to the surface and strobe them on and off. A wet surface will have a different angular variation in reflectivity vs a dry surface, unless the floor is very shiny.

1. https://thinklucid.com/tech-briefs/polarization-explained-so...

I'd like to know how the company behind the robot manages to get away with charging the cost of a small car for this glorified Roomba with a stick.

I must be in the wrong line of work or something...

Shoplifting aside, I am pretty sure those big funny eyes came after a meeting where they discussed how to keep the robot appearance on the right side of the uncanny valley.

This seems like a waste of money on the surface. The concerns over privacy seem a little silly given the hundreds of CCTV cameras that monitor the stores already, though.

Ok, this article is kind of all over the place and seems to miss several things worth mentioning.

1) It's pretty obvious that the responsibilities of the robots right now are just in a proof of concept phase. I can pretty much guarantee they will be used for additional tasks once the test period ends and some software bugs get worked out. Tasks like (as the article mentions) price checking, stock checking, misplaced item identification, etc. that take up a lot of employee time.

2) Customers concerned about privacy because the robots have cameras seem kind of out of touch. Ever been to Walmart and looked up? There are cameras everywhere, recording all the time and I haven't seen anyone complain that it's creepy they're being watched (and recorded) by people in a security office. The only difference with the robots is that they're more noticeable, so people are suddenly aware that they're being recorded.

3) The article mentions organized labor, employee opposition to the robots, and profits, but doesn't actually connect the dots. Here's the thing about capitalism and automation: the more you raise salaries and improve working conditions, the more incentive there is for companies to automate jobs. That's basic cost-benefit analysis and margin-based thinking. If automating a job costs X and human workers currently make Y, but want Y to be greater than X, guess what? Those jobs aren't going to last very long.

I think your second point is interesting, in completely practical terms I have to agree with you, you're already being recorded. Still, passive surveillance from a security camera feels very different from a robot with perceived agency actively following you around.

I dunno. In some aisles at Walmart, they make it really obvious that you're being recorded. The camera detects you and a ring of lights around it turns on. Greater theft deterrent, I guess.

> Greater theft deterrent, I guess.

The goal is to get you to look towards the source of the noise...which happens to be a lens aimed at your face. Now they have a clear photo of potential shoplifters.

#3 is missed by too many people. We need to start seriously considering what a "post-work" economy might look like, and we need to start drumming up the political will for the massive income redistribution that will be necessary.

I mean, employment is at an all timr low. People have been predicting a post-work economy for over a hundred years at this point. There’s not and never will be a post-work economy.

Not all employment is the same, especially the type that makes people feel like a cog in a machine, and pays similarly with poor future prospects.

We've been moving towards it for about as long as people have been predicting it. It's been happening in fits and starts, but it's been happening. I suspect that the rise of the "gig economy" is due in part to the fact that there is so much less necessary manpower required to keep society running. So the only work left is to cater to the desires of the rich and/or lazy.

> employment is at an all timr low

I think you meant unemployment here, but it's true the way you said it as well if you mean the labor force participation rate.

Reply on news.ycombinator.com