PITTSBURGH — Bossa Nova Robotics today announced its Bossa Nova 2020 robot for capturing inventory data and applying artificial intelligence to retail. The updated system includes a more nimble base for narrow aisles and onboard image processing, as well as lights and sensors to better scan products on shelves.
As brick-and-mortar retailers compete with e-commerce, they need to efficiently manage inventory, free up human resources for customer service, and provide a positive shopping experience for consumers, according to market analysts. Mobile robots are essential to reducing the number of out-of-stock and misplaced items, said Bossa Nova.
In addition, the company said its technology can provide real-time insights on shopping patterns so that retailers can optimize the location and stocking of inventory.
Building Bossa Nova 2020
“We’ve been working on this robot for two and a half years, and we’re taking a lot of lessons learned from the first version,” said Sarjoun Skaff, chief technology officer of Bossa Nova. “For example, our current robots have optics that focus on a portion of the shelves to read bar codes, but the other part of the shelf isn’t visible. To capture the entire shelf, they might have to make multiple passes.”
“Now with dual lighting and a redesigned camera, our robots can read the entire depth of the shelf, front to back, in really brilliant focus,” he told The Robot Report. “The robot spends less time on on the floor.”
Bossa Nova’s robots include 2D and 3D vision, high-speed and precision optics, and bar-code readers. For Bossa Nova 2020, the company used a combination of off-the-shelf and custom components, said Skaff.
“As with everything, we tried to see what was available and not build everything ourselves,” he said. “While our current robots were made with off-the-shelf components, to fulfill specific functions, we had to build the optics ourselves, soup to nuts, in collaboration with partners who are good at building cameras. For the computation — another lesson learned — we have a full custom computer with an amazing edge processor.”
In addition, Bossa Nova 2020 includes integrated self-testing and monitoring to reduce interventions by associates, said the company.
“It’s really the coming together of all the experiences we’ve had in stores since 2013,” said Skaff. “It’s a sign of the maturation of the solution that we can now build the right product for the market.”
Imaging on the edge
What about 5G networks? “We’re looking into it,” Skaff replied. “We think it will be trailing installations [of Bossa Nova 2020]. When it’s there, we’ll be ready, and we’re doing tests in the meantime.”
“In the PV20 design, now Bossa Nova 2020, we took the concept of the edge to its own edge,” he explained. “The custom processor can do all the computation we need in real time, and we also put compute in the camera. Images are processed onboard even before they leave the camera for final analysis in the cloud. We can read bar codes with full-resolution images on the camera and send the bar code out with a compressed version of the image, speeding things up. All the computation happens on the robots.”
“I don’t think there will be one approach,” said Skaff. “Stores will need the right mix at the end of the day. We’ve designed our robot for data capture in large-format stores, but we’ve architected the AI to be agnostic — it could take data from fixed cameras, drones, phones, etc. We’re figuring out the right combination of data capture, including mobile, cloud, and crowdsourced.”
While Bossa Nova is not licensing out its AI, it is open to providing services to other companies, he said.
“We’re thinking very actively about building interoperability for mixed-robot environments,” Skaff added. “We think the store of the future will have a myriad of robots from different companies for cleaning, inventory, security, and other functions, and they’ll need to coordinate their tasks. But what’s more, they could collaborate, exchange information, and alert one another, similar to the intelligent highway or connected cars. We’re very excited about that.”
Enhancing the shopping experience
While e-commerce has changed shopping habits and challenged conventional retailers on efficiency and cost, robots can help stores be competitive, said Skaff.
“Amazon this week announced yet another type of brick-and-mortar store in California, and if anyone knows what shopping trends are, it’s Amazon,” he observed. “Shoppers increasingly expect to buy anything they want, anytime they want, and that behavior includes brick-and-mortar stores. You want to touch an avocado before you buy it, see an assortment of products before making a decision, or talk to a human — there’s an experiential element of walking into a store.”
“Data is the central element,” Skaff added. “Bossa Nova 2020 can help a business know the state of a store and how to best allocate resources to fix problems, do more with less, and improve the experience.”
“In fact, a store owner has a competitive advantage in home delivery or buying online to pick up in store, because it has an installed base close to population centers,” he said. “Digital transformation breathes life into the thesis that physical stores not only need to exist but also have a future.”
Next steps for Bossa Nova
“The next step in understanding shopping behavior is an emphasis in online grocery,” said Skaff. “Produce has to be picked to taste — how ripe does a shopper want fruit to be? Unlocking online grocery is the next big wave.”
“Bossa Nova 2020 is modular, so a customer can add sensors for apparel or produce that we haven’t done yet,” he said. “We’re excited about the opportunities that it opens up with expanded coverage and providing access to stores with the robot’s smaller footprint, expanding our addressable market.”
“We’ll be showing the Bossa Nova 2020 for the first time at NRF in January,” Skaff said, referring to the National Retail Federation’s event in New York. “We’ve had a lot of inbound interest.”