How retail is using digital twins

Follow along with VentureBeat’s coverage from Nvidia’s GTC 2022 event >>

Digital twins began as a way to harness engineering simulation to improve product design. The Omniverse opens digital twins tools and techniques for a much broader set of use cases. At Nvidia’s recent GTC conference, executives from Lowes and Kroger explained how digital twins are transforming retail, customer experience and logistics. 

The biggest takeaway is how digital twins make it easy to visualize complex relationships between physical things, including product placement, physical customer journeys and the paths robots might take down store aisles for inventory and floor cleaning. 

Managers and staff can explore how layouts, schedules, team movements and customer journeys interact in one visualization tool. They can also visually assess the impact of a new store layout, schedule or technology might impact cleaning, restocking and staffing requirements. 

Digital twins also have the potential to improve customer experiences in various ways. They could help customers connect the dots between home improvement projects, required materials and materials costs. They could also help improve physical customer journeys within stores by organizing the order shopping lists to line up with a route through the store. 

Supercharging 3D at Lowe’s

Cheryl Friedman, vice president of Lowe’s Innovation Labs, described the company’s work on a new iOS app that automatically captures data about a home to simplify home improvement projects. 

Homeowners and contractors can quickly measure a room and the app can estimate the number of materials and the cost for different options. It can also prompt homeowners to find out the specs they may need when replacing appliances to minimize return trips. Consumers can also compare the cost of different paints, windows or doors vary across approaches. 

“There are few DIY projects you can get through without requiring some piece of data on your home,” Friedman said. “All of that information introduces some friction for home improvement journeys.”

Emerging tech like digital twins, mixed reality and computer vision help capture data about the home and keep track of all the details to reduce this friction. The Lowe’s app takes advantage of the lidar built into the latest iPhones to capture home measurements quickly. 

Lowe’s tried working with 3D and simulation tools using various commercial and proprietary tools in the past. But this introduced friction when employees wanted to move data across departments. “We are excited about Omniverse embracing Pixar’s USD because it creates an ecosystem that is open to all and this will supercharge our ability to use 3D at Lowe’s,” Friedman said. 

Omniverse also helps bring all 3D data together in one place to look at relationships across previously siloed data. This makes it easy to compare planograms that describe desired store layouts with how stores are actually laid out. Down the road, she hopes more simulation will help them improve planning for heavy store traffic on black Fridays or optimize scheduling for robots that clean floors and inventory shelves. 

Keeping it fresh at Kroger

Kroger is exploring how digital twins can improve product freshness and optimize logistics. Kroger is constructing digital twin models that reflect product freshness, meat cutting schedules and physical product layouts. The goal is to ensure all these moving parts go smoothly so that the lines are short, the meat cuts and produce are always fresh and the warehouses are well organized. 

Wesley Rhodes, vice president of technology transformation and R&D at Kroger, said, “It is like a dance you have to practice ten times and then you get it right. Technology makes something complicated look easy.”

Kroger has been experimenting with various schedule optimization tools for some time. But these were complicated for front-line workers to understand quickly. The visual nature of the latest digital twins built-in Omniverse helps workers and managers quickly visualize tradeoffs and get feedback when things don’t go as planned. For example, they could see how a new layout might cause traffic jams. 

Rhodes has found it helpful to borrow techniques from the airline industry. For example, airlines use video analytics to analyze the boarding process. New computer vision tools could similarly capture data about existing processes to reflect actual operations. This could help recalibrate the assumptions built into digital twins to increase simulation accuracy.

Better packaging

Down the road, digital twins could also improve product packaging, said Richard Kerris, vice president of Omniverse Developer Platform at Nvidia. A product package is the first thing a consumer sees when they buy a product. But package designers have always struggled with understanding how a container might look after it is dropped, under different lighting, or when stacked on shelves with competitive products. 

“You can get a true to reality duplicate of product packaging you are designing and look at it in different contexts,” said Kerris. 

For example, designers could simulate how different materials will look in the shop or how a package will behave when picked up or dropped. Designers can also simulate how different package design could affect box packing arrangements or individual shipping. It might be best packed and shipped efficiently and safely. Digital twins of the product package help simulate and visualize how a product package might look after being stacked in a shopping bag with many other products on the journey to a customer’s home. 

“You don’t want something that could break or crush,” Kerris said.