Computer vision brings intelligence to retail technology

From arrival to departure, the average time a consumer spends in a supermarket is approx 41 minutes for one trip. But when checkout lines are long and shoppers spend time scouring shelves for out-of-stock items, the journey can quickly become much longer. Neither consumers, who are quickly losing their temper, nor retailers, who are already dealing with post-pandemic staff shortages, supply chain disruptions and reduced foot traffic, want that.

That’s where cashierless checkout and inventory management come in, powered by artificial intelligence (AI) and computer vision. Different companies, both big tech and startups, have taken different approaches in recent years, using cameras and sensors to identify items and make phone calls – allowing the customer to quickly grab items off the shelf and leave without queuing. to stand.

Today, even as the economy slows, investors are showing no signs of pulling back investment in this sector. Major rounds of funding are still in the news, including Tel Aviv-based Trigo, which last week launched a $100 million series C investment, bringing total funding to about $199 million, according to crunch base.

Amazon Go has spawned many competitors

Six years ago, everyone thought that Amazon, with its novel Amazon Go, had the checkout-less supermarket model in the bag. But while many predicted the technology would scale massively, there are still only 28 Amazon Go stores around the world – and that includes the recent expansion in India.

In general, scaling is the biggest challenge, says Brad Jashinsky, marketing analyst at Gartner. “The main technical challenge behind the computer vision of POS systems is scaling the systems to handle more customers and larger store sizes,” he said. “There is also a need to integrate them more easily into existing store footprints.”

While Amazon has hesitated, several startups have stepped in with their own computer vision-powered technology – focused on solving scale and ease of integration. Some startups, like Caper AI – what was acquired by Instacart in 2021 – and Mashgin have focused on AI-powered plug-and-play smart carts or kiosks. Others, such as AiFi and Trigo, focus on ceiling cameras, shelf sensors and digital twin technology. While the economy is in a downturn, it appears that the opportunities in this niche sector are not being phased out.

Bringing computer vision to the physical store

According to Michael Gabay, CEO of Trigo, the appeal of computer vision to supermarket chains is winning over the capabilities currently available to them in an e-commerce space and bringing that intelligence into their physical retail spaces.

“That is a much larger market for supermarkets than the online market,” says Gabay. “Their Trust” [in technology] is much higher than last year or two years ago.”

In support of that statement, a Gartner study released late last year found that: 73% of retail respondents expect more investment in store technology before 2022.

In addition, retailers have been forced to refocus on connecting stores to their entire ecosystem, Jashinsky explains in an email. “New attention and investment has been given to the digitization of retail technology investments, including computer vision for cash registers,” he said.

Trigo was founded in 2017 by brothers Michael and Daniel Gabay, who grew up on a kibbutz in northern Israel and served in technological positions in elite units of the Israel Defense Force. The CEO and CTO respectively wanted to be the antidote to the headaches of waiting in long lines at the checkout.

The 3D store mapping and computer vision capabilities use artificial intelligence (AI) and machine learning (ML) to track a customer’s store tab as they go – even updating the total if they put an item back – and charge them accordingly when they run out, no lines are needed. It also tracks item inventory and customer body language so it can alert store associates if it suspects an item is hidden in a jacket, for example. All this without collecting biometrics or facial recognition data.

While that may sound like a lot for retailers to wrap their brains in, all they need to do is work with Trigo to set it up, and once the setup is complete – which is usually overnight, the company says – nothing more and they have the green light to start using it.

Despite the large number of competitors and the growing interest in the market as a whole, Gabay is confident in Trigo’s ability to stand out from the crowd. The company is currently deployed in supermarket chains worldwide including the Wakefern Cooperative in the US, the Tesco chain in the UK, the Shufersal stores in Israel, Aldi Nord in the Netherlands and REWE, a chain in Germany.

“We are the only startup and the only company to convert existing stores into standalone stores,” Gabay told VentureBeat. “We are also the only startup, not Amazon, currently targeting supermarkets and not just convenience stores or small stores.”

Trigo’s technology can be implemented by stores from 3,000 square feet to 5,000 square feet – and claims it is working on use in Stores of 10,000 square meters The next. But his competitor AiFi also claims it can be used in spaces up to 10,000 square feet.

Retail-oriented increase in computer vision

In particular, experts expect the computer vision market to rise sharply worldwide $41 billion by 2030. Investing in technology like this is the logical “next step for the industry,” McKinsey analysts Tyler Harris, Alexandra Kuzmanovic and Jaya Pandrangi recently wrote.

“Investing in technology used to feel optional for grocers – an opportunity to experiment or increase the ‘wow’ factor in stores rather than supporting mission-critical operations,” the article reads. “Today, a wide range of affordable, field-tested technologies can help retailers reduce the cost structure of their stores while providing a better experience for both consumers and employees.”

In contrast to the tightening of the current financial market, a survey by Battery Ventures found that 54% of C-suite executives have plans to increase their tech budgets next year – 75% say they have plans to increase it in the coming year. years at the very least. five years.

With its range of use cases from labor assignment, checkout convenience, merchandising, inventory management, loss prevention and maintenance – Jashinsky doesn’t expect the boom in retail computer vision innovation to fade anytime soon.

“The ability to provide ambient customer transactions is just one use case for smart computer vision technology for cash registers,” he said.

The real strength, he explained, comes from additional real-time business insights: “Retailers using computer vision for smart checkout can use the real-time insights captured to improve decisions.”