Sony Innovation Exec on Finding Use Cases for New AI Tech

By Meghan Hall |  February 7, 2023

A normal part of life for many: signs in restrooms that read, “Employees must wash hands.” 

But how might employers actually enforce that? 

New IoT-enabled motion sensor technology from Sony can help. It can count how many people are in the room at once, and can track their movements to see whether they washed their hands (and for how long), said Mark Hanson, Vice President of Technology and Business Innovation at Sony.

And while the sensor resembles a small camera, those who can access the feed don’t see actual people on the feed when they view it; they see anonymized, avatar-like figures, instead. 

The technology has a number of use cases — both to track the movement of people and products in the supply chain. Sony has built AITRIOS, its edge AI platform, in partnership with Microsoft, which is acting as the cloud partner. The technology can enable efficient shelf restocking by detecting specific, quantifiable gaps in stock, or track how many people are in a space at any given time to manage crowds, or understand places where customers are making decision points inside of a store.  

Hanson said, in the case of shelf restocking, AITRIOS can increase frontline employee efficiency, an issue many leadership teams have been focusing on in the past several years. Employees are told exactly which SKU needs to be replenished; how many of that item they need to bring; and where the item is located on the shelf. 

NRF conference-goers get a glimpse into a use case of Sony and Microsoft’s new technology — live shelf monitoring for efficient restocking throughout the day.

Hanson’s Team and Directive

Hanson works on Sony’s Semiconductor Solutions team, based in San Jose, California. 

Mark Hanson, Vice President of Technology & Business Innovation, Sony

“My responsibility is two-fold,” he explained. “One, I do a lot of strategic partnership type activities. And [I am] kind of the evangelist of technology. I’m also in charge of marketing.” 

Hanson said his team functions as “a startup within Sony.” It uses internal accelerator programs to stimulate technology exploration and growth, then works to get its activities funded to scale. 

“We’re very experimental. Getting to the business side of it, we have to put a business case together, get the [leadership] to approve that it’s worthwhile to do it,” he said. “When you’re building in Silicon [Valley], it’s a pretty big investment,” given the salaries that talented engineers and product managers command.

The Process of Building the Platform

AI entered the hype cycle in recent months, with the explosion of so-called generative technologies like chatGPT and DALL-E captivating users. But artificial intelligence has been prevalent for some time now; it’s used to power self-driving features in cars; help personalize mobile phone experiences; and automate processes in manufacturing plants. 

Hanson said this development has been in the works for a few years — and that it started with an idea he called “crazy.” 

 That idea resulted in combining an image sensor and a logic chip — which together process information to complete tasks — and putting that combined chip inside a small camera. Thus far, Sony has only released one model: the Lucid Sensaiz. It requires PoE (power over ethernet) to run. Two of the sample models Sony showed at the National Retail Federation’s 2023 conference will be able to run on battery or USB power in the future. 

And while the technology is in its early release stages, having made its debut in January, Hanson said Sony now has demand to prove further expansion will be worthwhile. 

“How long will it take to make this a business? We’re at the infancy… so I don’t know the answer to that, but we’re hoping it’s soon,” he said. “The good thing right now is that [we have] lots of interest in trying to figure out how to get to the scale level… We’re not necessarily interested in just selling a chip. We’re interested in creating an ecosystem — something that accelerates the development of edge AI today.”

While the idea to combine image sensors and logic chips seemed like a moonshot at the time, over the course of several years, that foundation has turned into Sony Semiconductor Solutions’ first platform-as-a-service model. 

Hanson shows how AITRIOS converts real-time images into anonymized, avatar-like figures where humans are involved.

Collaborating with Developers to Identify Use Cases

Hanson said Sony will help developers build use cases for this technology every chance it gets. 

“Anybody who’s got a customer…[or] who has a proof case that they’d like to try with this kind of technology, we’d like to work with them,” he said. “There’s no cost. Any of those developers — or end-customers, if they’ve got their own technology — could come in and… we would [say], ‘Tell us what the use case is. Let’s figure out what the data is that you need to make the use case.’”

Hanson said, from there, the developers can begin building the use case, which they own, despite it being built with Sony sensors and cameras. 

“They can build their own use case [using AITRIOS and Microsoft Azure] and proof it. After that, we can make agreements on extending it, expanding it, or commercializing the solution.”

Photos courtesy of Sony Electronics.