According to The Asahi Shimbun, summer spells fun for many, but not for air conditioner repair workers, who face incessant calls during the peak service period and often have to make repeated visits to fix a single problem.
And since the average air conditioner has 2,000 or so components embedded inside, it's no easy task to figure out which parts they need to take with them when heading out on a repair call.
To simplify life for those sweating to keep customers cool, leading air conditioner manufacturer Daikin Industries Ltd., based here, is relying on an artificial intelligence (AI) system to pick the parts.
Few other firms in Japan are utilizing an AI system on such a large scale for repairs, Daikin officials said.
As summer nears, the company's call center in Osaka’s Chuo Ward begins being bombarded with calls from households and businesses requesting repairs. Callers typically complain that their air conditioner is emitting an odor or isn't effectively cooling their room.
In response, Daikin’s service centers across Japan arrange necessary replacement parts and assign repair personnel to visit customers. The company gets 2,000 to 3,000 repair requests on average every summer, Daikin officials said.
Repair workers rely on experience to decide which parts to take with them, but often realize once they are at a job site that some necessary parts are missing.
Only 53 percent of all repairs Daikin did last summer that required component parts were wrapped up with a single visit.
To reduce parts mismatches, Daikin introduced an AI system from Abeja Inc., a Tokyo-based business venture. In a trial, the system was fed 10 years of data on repairs and made the right selection in 64 percent of all repair requests that Daikin received last summer.
Additional learning by the system is expected to further improve its accuracy, so Daikin has decided to leave all the selection work to it starting from June.
Fewer mismatches are expected to also cut expenses. It costs about 700 yen ($6.49) to return every single unnecessary part to a warehouse, officials explained.
Daikin also plans to rely on an AI system to assign tasks to repair workers in accordance with their skills, the officials added.
“Although there are calls for more use of AI systems, few workers in manufacturing industries have any know-how on how to use them,” said Hitoshi Yamamoto, head of planning with Daikin’s services headquarters. “It remains quite uncommon to have an AI system learn how to select component parts and assess what skills engineers have. We hope to remain the industry leader into the future.”
If you want to read this article in Japanese, please see the following link:
Subscribe to our English Newsletter