Greg Buzek is founder and principal of IHL Group, a global research and advisory firm specializing in retail and hospitality technology. He has more than two decades of experience in retail market analysis, business planning, product development and consulting with Fortune 500 companies. He spoke with NRF about COVID-19’s impact on robotics and other machine-oriented retail technologies.
How has COVID influenced the use of robotics and other high-tech tools?
With COVID, there’s a long list of technologies sold as labor-saving that have been reframed as customer safety technologies. As a result, we’re seeing a massive increase in the use and purchase of those technologies.
Check out more of the latest information that we have on retail tech here.
We’re in a race to get to accurate inventory. The first one there wins. That includes computer vision, robotics, automated warehouse, automated pick and RFID. It also means computer-aided ordering. We saw situations where the computer said, “Order 40 times more than you typically order of things like hand sanitizer.” A human being came in and said, “That can’t be right.” And they lost out on those sales because the order was cancelled.
[Some companies] are racing to deploy these technologies. There are others who have their heads in the sand, struggling for survival and not able to do it. We are seeing some radical separation between the haves and the have nots that will define the next decade or more.
What will be the long-lasting impact of COVID on machine deployment?
At the end of the day, it’s about having the inventory to meet customer need. We had $1.8 trillion of out-of-stocks and overstocks this year. We had $500 billion in out-of-stocks that happened to be because of COVID. A lot of those were previously hidden.
When we walk through the stores, we’ll make an alternative decision when we encounter an empty shelf, and the store won’t even know. Yet now the number of times that you’re dissatisfied as a shopper has gone up dramatically. Those who will win in the future will be the ones that have what you want when you want to buy.
In the United States alone, we spend about $24 billion on safety stock because we don’t know what that accurate inventory number is, just hoping that the safety stock will cover us.
What are some aspects of robotic development/technological innovations that you’re watching?
Fully automated warehouses have great potential, particularly in going into inner-city locations, going vertical. There are times those work, particularly where it gets harder and harder to get people who want to work those jobs. When it comes to warehouse jobs, there’s great competition to get people there and those wages continue to go way up compared with traditional retail jobs.
Because of that, the labor starts to price itself out and the automation prices itself in. If it’s a 5 percent increase in wages on an annual basis, when it comes to technology, we’re seeing a 5 to 10 percent drop in costs every year. We’re getting to that spot where full or at least partial automation makes sense.
If we were having the conversation a year from now, where would robotics be thanks to the groundwork laid during this period?
I don’t think it’s going to change that much in terms of new technologies. It’s more the level of use of these existing technologies. Where we’re seeing the biggest growth is in technologies that have been around for a while, but they now have tacit approval to use. Self-checkout is 20 years in the making and Walmart has just said they’re going to go full self-checkout in a trial. C-stores are going to self-checkout at the peaks. It’s not like it hasn’t been there and installed. It just went from an install rate of 5 percent per year to 15 percent.
We’re getting tremendous gains with computer vision and some new enhancements that will happen with RFID. New versions of that start to incorporate LiDAR (light detection and ranging) which penetrates items. The limitation on computer vision installed most places is that it can only see the front of the shelf and not beyond that. It can’t tell you how many items there are.
When you get LiDAR — now on the iPhone 12, which will dramatically drop the price — computer vision is going to be much more accurate on telling you how many items are on the shelf. The full solution is going to be some combination of RFID and LiDAR-enhanced computer vision.
Edge computing, where data is stored closer to where it is needed, will be a big conversation as well. Is that information best computed at the shelf, in the actual robot, in a server or up in the cloud? We’re seeing 580 percent increase in the next two years for edge computing. All these IoT devices limit your local area network ability in the store.
The one thing that has really come out of COVID is just how you can’t afford to have a single point of failure. You can’t afford to have one of anything, except one view of your inventory levels and customer data. All other systems will need redundancy of solutions, from sourcing products to networks to supply chain. If you go to a fully automated warehouse and you have a ransomware attack, you’re shut down. People are much more aware of the fact that we have to have redundancy in all areas.