Earlier this year, we wrote about machine learning & image processing in embedded systems using pasta type classification as an example.
Lixil Group, a Japanese company manufacturing housing equipment, is working on a concept using similar technology, but instead of classifying pasta, they implemented the technology into an AI-based toilet that checks the shape and size of feces of people living in elderly facilities.
Lixil’s toilet was unveiled at the 46th International Home Care & Rehabilitation Exhibition (HCR 2019) in Tokyo at the end of last month. The system relies on a camera and two LED’s placed on the back of the seat, which turn on when the user sits on the seat. Once the job is done, images of the feces are taken automatically and classified into seven categories as specified by the Bristol stool chart.
Classification happens within a few seconds, and results can be check by staff members using a computer or tablet. The goal is probably to improve services, as staffers would then know if a patient has diarrhea for example.
You’ll probably be glad to know the AI toilet was tested on Lixil’s employees with 3,000 images, and judgment accuracy is now 80% or higher.
Lixil plans to conduct a verification test at an elderly facility in Q2 2020, but somehow Nikkei XTech reports the company is not planning to commercialize the system. Maybe that means the current prototype will not be manufactured, but eventually, the technology may be implemented in future AI-based toilets from the company.
If somehow you’d like to get your own smart toilet at home, to control it from your smartphone instead of the buttons in the unit, we previously covered Bluetooth LE controlled Kohler Novita therapy bidet.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.
Support CNX Software! Donate via cryptocurrencies, become a Patron on Patreon, or purchase goods on Amazon or Aliexpress
…with automatic upload to Fecesbook? /scnr
Shit, AI is worse than we thought!
This seems like a good application of AI. Japan has a massive looming issue with caring for the elderly. Presumably AI like this toilet allows more of their care to be provided by less trained carers like the ones the new immigration policies are meant to help source from aboard.
One product that’s actually on sale is an IoT water pot that lets a carer see that someone is following their daily routine without sticking cameras all over their house.
Ok, feaces recognition is not as bad as facial recognition as that can be fatal…
>facial recognition as that can be fatal…
Facial recognition has it’s uses too. I think that sort of tech will have a lot of good uses in screening people for medical conditions again using untrained/less trained staff. The issue is going to be when CCTV video is fed into the same tech and instead of booking you in for a check up it sends a health warning to insurance providers.
I’m wondering how long the camera will remain clean enough to take accurate pictures. Even if cleaned daily, water traces could significantly degrade the images. BTW, am I the only one feeling like the TV is proposing the menus at a restaurant ?
> BTW, am I the only one feeling like the TV is proposing the menus at a restaurant ?
Did you imagine a conveyor with little pieces of poo travelling around on dishes?
They’ll know when somebody has explosive diarrhea because the shitty thing stopped working…
I’ll take a Number 7 with extra sauce, please.
To think, only a matter of months ago we would have poo-pooed the idea of an AI scatologist. Now I expect it’ll become quite the regular thing.
How do they know which resident is delivering the sample?
Does the resident first have to ask permission to go to the bathroom in order to be identified?
Or are all the residents chipped, so that the analysis system can identify who is who when then use the facilities?
[Being chipped is not just for animals and children any longer …]
Weight might be enough to identify each user.
poopsplash proof? 🙂