The Art of Automation: Chapter 10 - Automation in Retail

11 min read

A conversation between Jerry Cuomo and Tim Vanderham from NCR Corporation on automation in the retail industry.

aoa10

Covered in this chapter

  • Automation in the retail world
  • Retail on the cutting "edge"
  • AI in the retail world
  • Crypto in retail

The Art of Automation: Table of Contents

An industry lens

This chapter of the Art of Automation is a reduced transcript of a conversation between Jerry Cuomo and Tim Vanderham, Senior Vice President and Chief Technology Officer at NCR Corporation. In our third chapter of AI-powered Automation viewed through the lens of a specific industry, Tim and Jerry discuss how progressive retailers are beginning to use automation to transform the customer's shopping experience — a transformation that has been catalyzed by the ongoing pandemic. 

Tim shares automation examples from self-checkout to online ordering and explains how this goes beyond classic automation to edge, computer vision, Internet of Things (IoT) and even cryptocurrency. They close by jumping the tracks to artificial intelligence (AI) and diving into how, out of necessity, retail companies are utilizing AI to filter through millions of events to isolate outages and create a more personalized experience for their consumers.

Jerry and Tim 

CUOMO: Welcome everyone to the Art of Automation, a podcast that explores the application of automation in the enterprise. Folks, today's "always-on" world has led to new rules for engaging consumers. As shoppers' needs and demands change, retailers need to respond accordingly and quickly.

Consumers are increasingly expecting efficient, safe and engaging online experiences to also be replicated in store; and as you'll hear from my guest today, progressive retailers are using automation and AI to transform the retail experience — making sense of immense volumes of data, filtering out the noise from millions of events generated in and above the retail store. They're using advanced technology like edge computing, Internet of Things, computer vision, hybrid cloud and even cryptocurrency.

So, for today's episode, we have the fortune to have my good friend Tim Vanderham with us. Tim is the Senior VP and Chief Technology Officer at NCR Corporation, leading NCR's global software and technology organization.

His team includes software innovation and software engineering, with more than 3,000 engineers building NCR's current and next-generation products and solutions. With that, I'd like to welcome Tim to the Art of Automation.

VANDERHAM: Hey, Jerry, thanks for having me. It's great to be here with you today.

Automation in the retail world

CUOMO: I'm so excited, Tim. I want to get right into it. Why are you so excited about the opportunities to transform the retail space; and, why now? What is happening in the world today, and why is that putting an exclamation point on all of this?

VANDERHAM: Yes. So, you know, here at NCR, the passion of automating retail is really out of necessity. When you think about what retailers are going through in the world today with rising labor costs, the challenge they're having of keeping their shelves stocked with the demands that are coming to people in the store, above store ordering, etc. — automation is the way they're going to still make the bottom line and still serve all of us as consumers.

And COVID accelerated it. We saw this as a necessary coming down the pipeline — we were planning for it.  But COVID hit March of 2020, and bam, we're in the middle of — "how do we better automate the retail experience and the retail store."

So, I'm excited about it, my team's excited about it. And more than being excited about it, our customers are demanding it. They're coming to us asking: 

  • How can we deploy software solutions faster?  
  • How can we automate inventory management?  
  • How can we automate that whole notion of, "I order online, I pick up in the store"?
  • How do we automate that flow end-to-end? 

Now think about it in the convenience store space, think about it  in a department store like Macy's. It's happening in spades across all of our retailers, across the globe. It's exciting.

CUOMO: Yes, I can only imagine. I mean, I'm a consumer — I'm part of this on the other end — so I see what's happening and really appreciate it. But tell us a little bit more about some of the processes that you're focused on. NCR is known for the cash register, but you're so much more than that these days.  Share with us some of the processes that are working behind the scenes that are targets for automation.

VANDERHAM: Sure. So, how are we running the store more efficiently? It's everything from self-checkout, which we have market share leadership in when it comes to self-checkout machines across the globe. But also then how do you just manage all of those components that run in your store?

When you go into a Whole Foods (let's use as an example, a great customer of ours), every self-checkout lane or every manned checkout lane is literally running a computer today, right? Running a piece of software that you interact with as a cashier or as a consumer in self-checkout. How do we keep that up and running? How do we keep that at an SLA level available every day, every hour of that day? And so in-store automation is really key.

We automate things in the store, but then we're using IoT technology to be able to get alerts off of it so that if one of the processes goes down or a printer goes offline we can restart through automation, through advanced analytics and automation — restart that printer, restart that process so that manned lane or that self-checkout lane is only down for a couple of minutes versus having a store manager call, somebody log in, somebody to drive a truck to fix a printer, whatever it might be. So, that's kind of one example of automation in the store.

CUOMO: Yes, okay. And Tim, just give me a sense here — I can imagine years ago where the cash register was a physical device that had probably firmware that got updated infrequently. What is your frequency of change? Is it measured in weeks, days or finer-grained than that? How often are you automating changes at that level?

VANDERHAM: Yes, so, it has been increased. It's not where it needs to be yet today; I hate to say it. You know, I manage everything from code that's been built in the last 12 months to code that's literally been built 25 years ago. And so, that frequency of change depends on the code level and the solution that's running in the store.

We're trying to get it down to in the order of updates every month when it comes to kind of core software packages. And down to literally every 10, 15, 20 minutes when it comes to alerts coming off of the devices and us taking automated remote actions on those. So, it's a wide spectrum. And that's just in the store, and then we've got to think about above the store.

CUOMO: Yes, tell us more.

VANDERHAM: We started thinking about online ordering. When I order online, it sends in the solution, somebody goes and picks, oh, the steak that I want is out today. You have to be able to real-time interact with the consumer; again, through an automated way, using the picker device in the store and me through my texting or chat app that I'm using on my mobile device. And so, we're automating that. The above-store stuff we are literally updating, you know, multiple times a week. You see this paradigm of a hybrid environment. We've been talking about hybrid in software for a lot of years.

CUOMO: Sure, sure.

VANDERHAM: This hybrid environment, what runs in the cloud, what runs in the store and then how we manage those, I'll say homogeneously, but on different intervals because of still some of the complexity that comes. And then maybe during the session, we can talk about some of the things we're doing at the edge to made edge look more like cloud in the long run.

Retail on the cutting "edge"

CUOMO: Let's jump right into that. Let's talk about edge. What's the role of edge here, and what's running at the edge? What's the interaction between the edge and the mothership and how does automation play a role in that?

VANDERHAM: Yes, so, about three years ago now, we acquired a company called Zynstra where we started virtualizing at the edge. So, we could take those old Windows monolithic code bases and at least virtualize them so that, again, you can shrink wrap the VM, test it, configure it and deploy that and do almost near time — no downtime updates. So,  we're running a lot of that at large retailers today.

One of our great examples is Pilot Flying J. So, over 500 locations for truck stops — everything in their store runs virtualized in the back of the house with a dumb thin client on the front end. They're kind of on the leading edge of this virtualization. So, that's what we've been doing for the last three years. And now the team's working on — as we break apart those monoliths into microservices — how do we run that application or that set of services, set of components at the edge in a container-based model?

We're literally bringing down on that same infrastructure MicroK8s infrastructure, running sets of microservices. So, think of a selling engine, think of a tax engine, think of an inventory engine and then all those run in store in the back of the servers with a lightweight thin client. And then we're going to be able to update those in not a monthly cadence but in a daily cadence and manage SLAs at the edge.

AI in the retail world

CUOMO: Oh, very cool, Tim.

Hey, Tim, let's jump to tracks here. This is the Art of Automation — we talk a lot about the relationship of AI and automation. So, could you add the AI perspective to the story here? What's the role of AI in automating retail?

VANDERHAM: AI serves a huge role because we have to be able to take all of the sensor data off of the hardware, the software — even human interactions. When you start thinking about how we can instrument our mobile devices or our websites, we can start to see people's activities, right? And so leveraging that super set of data — again, machine data, software data, transactional data (which is also really important) and then consumer behavior data.

Bringing all that data back to a common data lake, leveraging AI and advanced machine learning algorithms against that to make sure that we are doing the right things for our customers. We're committing, NCR is committing to our customers an SLA — so, a Service Level Agreement of when your systems are up and running, able to transact and that you're able to transact 24 hours a day, seven days a week in many cases.

We have to use AI to intelligently filter out the noise so we can focus on what really needs to be fixed, because let's be honest, some of these are machines. Sometimes we have literally filter out the noise, get the telemetry data back and say, we've got three self-checkouts down out of six at this one store for various reasons. Get a customer engineer in a truck rolling to that store to get them back up and online for when they open at 6:00 a.m.

So, that's the level of sophistication that we're working through of hundreds of thousands of events every day from every retailer, so millions across our retail set. How do we filter it out, how do we give support and, ultimately, how do we deliver availability. So Jerry, when you're a consumer and you want to walk in to Whole Foods or you want to get your gas at Sheets there in North Carolina — both good customers of ours — you're able to transact because our systems have made sure we're available.

CUOMO: I've seen a situation where I was about to scan a bottle of wine, and it could have been my imagination, but it seemed like the checkout manager got an alert to ask for my license before I even kind of pulled it out of my cart.

So, are we in a world where there's that level of granularity of awareness in the store? Is AI — things like computer vision — is it really an alternate employee now in the store? Maybe paint a picture of where this is going. What's a day in the life of a consumer going forward?

VANDERHAM: Yes, you're absolutely right. We call your self-checkout environments "bullpens." When you think about all the cameras in the bullpen — yes, there's a camera in the scanning agent in the self checkout machine. Many of them have a camera above head now as well to avoid shrink, and that could be what's picking up objects that are in your basket.

So, this isn't widespread yet, but when you start thinking about that, yes, the eyes in the sky around those bullpens are going to help us understand when baskets are really empty and when they're not so they can avoid shrink. They're going to hopefully help identify you, Jerry, as who you are, if you're a loyal member.

So, some of our retailers are wanting facial recognition so that they know that, "hey, Jerry, you're back in today, thanks for your business. Oh, by the way, we're going to give you a special discount on something that will show up in your mobile device as well just because we came and talked to you today or what have you."

So, really that whole heatmap, tracking, understanding. And at the end of the day, we all want a better consumer experience, so we're going to leverage computer vision, facial recognition, advanced AI concepts if you opt in, obviously, from a data security perspective, but if you opt in, to give you a better experience, which also makes that consumer better able to serve you day in and day out.

Crypto in retail

CUOMO: On that topic of painting a picture of the day in the life of a customer, you've been in the news recently for your acquisition of LibertyX and I think being a very progressive point of sales organization thinking about crypto. Can you quickly comment on some of your ideas around crypto and what you're thinking?

VANDERHAM: Yes, absolutely, Jerry. So, you're right, we acquired a company called LibertyX, they're a crypto-based company for three things. So, we can allow consumers to buy and sell crypto in a digital format or with cash at a physical device like a self-checkout machine or a point of sale. So, we can help consumers using crypto as the rails which is a more consumer friendly way and a safer way to move money from, say, to the U.S. to Brazil or U.S. to Mexico.

Finally, it's payments. And so, I want to allow consumers to be able to take whether it be Bitcoin or other alt coins or a stable coin like USDC, and whatever you have in your wallet, if a consumer wants to interact that way, we have to enable that at that point of sale at that self-checkout machine.

So, that's kind of our vision of what we're doing, and I like to say we want to put crypto on every corner, because we have hundreds of millions of touchpoints every day digitally and physically with consumers. So, Jerry, next time you want to buy your groceries or your gas with Bitcoin or other cryptos in the future, look for an NCR pump or an NCR point of sale and you'll be able to do it.

CUOMO: Will do, Tim. Thank you so much, that has both been inspiring, impressive and so progressive. Thank you so much for spending time with us this afternoon.

VANDERHAM: Thanks, Jerry. Great to be here, and thanks for the time. Always great to see you.

Learn more

Make sure you check out The Art of Automation podcast, especially Episode 21, from which this chapter came.

Make sure you check out The Art of Automation podcast, especially Episode 21, from which this chapter came.

The Art of Automation: Landing Page

Be the first to hear about news, product updates, and innovation from IBM Cloud