find what they have lost

Losing something in public transport is a more common situation than it seems. On many occasions, returning to the place where it is believed to have forgotten is not enough. And in a country with both movement and Japan, the challenge is even greater.

A platform that wants to put order in chaos. To face this daily problem, Japan has begun to display a technological platform called Find, based on artificial intelligence (AI). As the firm itself saysit is already operational in train stations, airports and other public spaces in the country.

Its operation is direct. When someone delivers an object lost in one of the enabled points, the staff takes a picture and introduces it into the system. Find analyzes the image, identifies parameters such as the color, the shape or distinctive details, and stores all that information in an accessible database.

Visual recognition, messaging and automatic search. The next step is the user who has lost his belonging. Through applications such as Line, tremendously popular in Japan, you can start the search indicating where you think it lost it, a description of the object and, if you have it, an image. In case of not having a photo, the system allows you to select a similar visual reference.

With that data, the AI ​​promoted by GPT-4O Openai performs an automated search and returns a list of possible coincidences ordered by degree of similarity. Even if the description is inaccurate, the system manages to generate relevant results. If it is finally confirmed that the object belongs to the user, it is indicated where you can go to pick it up.

The service is available in Japanese, English, Chinese and Korean, which makes it an accessible tool for both residents and foreign visitors.

Of spreadsheets to automated analysis. According to SCMPthe technological jump has been remarkable in some operators. Yurikamome, which manages a monorraíl between Shimbei and Toyosu, in the coastal area of ​​Tokyo, used until recently spreadsheets to record the lost objects. A completely manual system, without any automation.

Since July 2024, with the incorporation of Find, the process has been transformed completely.

A measurable improvement from the first moment. The Keio operator, based in Tokyo, has been one of the first to quantify the impact. Before incorporating the new system, just 10% of lost objects returned to their owners. Now, the figure is around 30%. A remarkable improvement that suggests that the solution works, even in this first stage of deployment.

Technology to solve real problems. Although there are similar initiatives in airports or airlines, the Japanese approach stands out for its technological integration. Combining image analysis, instant messaging and automatic learning, Find proposes a new and effective way to manage a daily situation.

Images | Find

In Xataka | If Apple does not want to be left behind in the AI ​​race, you have a very obvious option at your fingertips: Buy Anthropic