How a Neural Network Runs a Family-Owned Japanese Dry Cleaner

The street finds new uses for corporate technology.

Closed Circuit

In the Japanese city of Tagawa, automation could one day enable dry-cleaning businesses to operate without any employees at all.

At least that’s Daisuke Tahara’s goal, according to WIRED. Already the owner of eight dry cleaners, Tahara taught himself the basics of machine learning and used them to build a system that lets customers create a service ticket by laying all their dirty clothes on a table to be scanned by a computer vision system.

Trickle Down

Tahara’s story is one of four short profiles published by WIRED on Tuesday. The common thread is that each subject decided to tinker around with artificial intelligence, usually teaching themselves to harness an algorithm’s ability to solve an everyday problem or, in Tahara’s case, streamline his family’s business.

AI giants have strong incentives to keep all of their developments to themselves. But they often choose to share some of their research to help foster a stronger research community and, presumably, to boast.

If you Build it

These glimpses into top AI labs can help the tech-savvy get started with their own AI pet projects.

The most impressive AI developments are still likely to come from those tech giants that have an absurd amount of money to throw at engineers and their labs. But in this era in which most AI research is on new uses for algorithms rather than major breakthroughs, more diversity among those who can build AI will only help the technology move forward.

This story has been updated to reflect that Tagawa is a city in Japan, not a prefecture.

More on the AI community: An AI Conference Refusing a Name Change Highlights a Tech Industry Problem

Share This Article