We often write about artificial intelligence (AI) being used on food before it gets to you: inspecting the supply chain, making sure your burgers are cooked, etc. But a new unit of Japanese company Hitachi is applying AI to food leftover on the plate after people are done with it.

The Japan Times reports that Hitachi is partnering with hospitals to use AI to analyze food not eaten by inpatients. Hospitals prepare meals with specific nutrients for each patient to assist in the recovery during their stay, and uneaten food can result in recovery delays.

Hitachi has already starting testing its tech with a major hospital in Japan. The system works by using a camera mounted on a trolley that collects trays, taking pictures of the leftovers. The company’s deep learning algorithms then examine the images to provide analysis.

By doing this post-meal analysis, Hitachi’s systems can recognize patterns in the leftovers that humans otherwise could not see. Japan Times writes that nurses often check leftovers now, but the task adds to their workloads and they are not trained nutritionists.

Hitachi plans to use this system even for remote patients, who could take pictures of their food with their smartphone. If this system works as promised, it’s not hard to envision this technology moving further into the consumer category. Being able to quickly analyze your (or your kid’s) uneaten food and establish a nutritional pattern for what’s left would go a long way to help identify personalized food profiles, or help customize your own personal 3D printed food.

Of course, it seems like the end goal for this technology is self-obsolescence. The better the AI works at identifying food you haven’t eaten, and the better you’ll (hopefully) get at eating all of your food — the less you need the leftover scanner.

Leave a Reply