When Jordan Fisher was looking for a way to reinvent an aspect of the physical world he felt was long overdue for change, his gaze quickly turned to the checkout lane.
Fisher, a machine learning expert who is the founder and CEO of autonomous checkout firm Standard AI, believes traditional computer technology is nearing the end of its run as an agent of disruption, in large part because artificial intelligence has the power to bring about massive changes beyond the digital realm.
That realization led him to conclude that computers with the ability to interact with the physical world had the promise to finally revolutionize vast swaths of life that had long been stuck in the past.
“Traditional software’s only really good with digital information, and the vast majority of world and commerce happens in physical reality where traditional software just couldn’t make an impact” Fisher said in an interview. But with artificial intelligence in hand, “it only takes a couple of minutes before you start stumbling upon something like autonomous checkout, and you’re like, ‘OK, this is huge. Ninety percent retail is physical retail, there’s billions and billions of hours that people just spent waiting in line every year. Can we have an impact that was previously not possible?’”
Fisher is leading a company that is looking to use its expertise in machine learning and computer vision to eventually eliminate the need for humans to oversee the checkout process in retail locations. For now, Standard AI is focused on smaller retailers like convenience stores, but Fisher believes that the autonomous checkout tech his company and competitors like Amazon, Grabango, Zippin and Trigo have developed will become a common sight across the retail landscape, including in full-size grocery stores.
Fisher noted that he was also drawn to the car industry, which he believes is another example of where machine learning is playing a key role in pushing society ahead, but decided that experimenting with moving vehicles was just too risky.
“You can think of autonomous checkout as an inverted self-driving car, so the cameras are pointing outward instead of inward, and instead of the car moving, the people move,” Fisher said. “But we get this luxury of being able to make a mistake without killing anybody, which is really important.”
A mathematician who worked as a machine learning specialist for the U.S. Securities and Exchange Commission before founding Standard, Fisher recently spoke with sister publication Grocery Dive about the future of autonomous checkout.
This interview has been edited for length and clarity.
GROCERY DIVE: Autonomous checkout technology has been around for several years, but has been somewhat slow in showing up stores. What do you think is keeping the technology from faster adoption by retailers?
We picked this application space because we thought it’d be a little bit easier than self- driving cars. But it’s still been a really hard market to start growing. It’s a physical product that requires physical operations to install and to maintain. It’s not like software where it’s just you flip a switch and you can reach 100 million people the next day.
To what extent do you think cost is an obstacle?
We’ve been working aggressively to get the price down, and that’s a core requirement. A lot of retailers have a lot of cash flow, but they don’t really have a lot of that spare for capex on hardware. So getting this to work was step one, which we’ve done, showing that we can start scaling it and make this repeatable is step two, and we’ve done that as well. And now step three is we have to make this a really economical product for everyone involved.
When do you expect to bring your technology to retailers larger than convenience stores?
It’s hard to predict, more because the c-store market for us is huge, and we’ve got our hands full. I think you’re going to start seeing other innovations in grocery even before you get to full autonomous, like identifying produce at self-checkout stations, so we can make a better checkout experience. And I would expect to see that pretty quickly over the next couple of years.
Other than checkout, what other uses for computer vision do you see in the retail arena?
Part of our thesis is because you have a smart system that understands where things are in the store, where people are in the store and what they’re doing, you can use that for other things. We’re now actively working on getting other features out there, like out-of-stock protection, to help employees do their job more efficiently. I think working should mostly be fun, and we should try to find the parts of a job that suck and are kind of drudgery and let a machine do it, so that they can focus on the parts that are more engaging from a human perspective.
Standard recently acquired Skip Checkout, a company that is focused on traditional self-checkout technology. Why did you decide to purchase it?
Part of the strategy for us is that if you if you have a checkout need, you should be coming to Standard, whether you’re on the beginning of this path where you’re looking to do self-checkout, or you are trying to modernize your point-of-sale system, or you’re looking to save even further steps to getting to full autonomous. It doesn’t really matter where you’re at on the spectrum. You come to Standard, and we’ll have the right package that suits you and takes you on that voyage.
The self-checkout unit becomes the kiosk for the autonomous system when you decide to invest further in the checkout experience. So it’s kind of like an on-ramp.
Amazon has played a key role in introducing autonomous checkout technology to the public. What are your thoughts about the competition they pose to Standard?
I think it’s great to have Amazon in the space. Retail is a $25 billion market, and there’s room for all of us. I’ve always been encouraged by having them in the space because they help with things like shopper education.
I do think there’s places where they’ve made product choices that I wouldn’t have made, because they were originally thinking of this as a vertically integrated technology for themselves. But I’m rooting for them because I think it’s healthy for the industry to have another successful player.
How do retailers tend to respond when you make the pitch that they should consider installing an autonomous checkout system?
It’s very rare for us to talk to a retailer and they say they’re never going to do it. What’s usually the case is that if they’re not ready, they recognize the need for it.
What are your thoughts about the risks to privacy posed by artificial intelligence?
I think the new wave of generative AI that we’re seeing today has all these potential for abuses in terms of disinformation, an example. I think it’s incumbent upon us to be deploying these things in places where that’s not going to happen.
How do you respond to people who are worried about being tracked when using your technology?
We’ve always held privacy and trust as key tenets of our product. We built it in from day one. We don’t do facial recognition, even though building a system like this is easier if you do things like that, because it makes it easier to track a person across the store. But we don’t do that. We have no ability to recognize people digitally, and we’ve intentionally not built any of the components of our system to allow for that.
This article was written by Sam Silverstein from C-Store Dive and was legally licensed through the DiveMarketplace by Industry Dive. Please direct all licensing questions to [email protected].
