Imagine losing your phone on the subway in a big city. The chances of getting it back are slim, right? In Japan, where public transportation moves 12 billion passengers a year, this problem was a logistical nightmare.
Until recently, only 1 in 10 lost items returned to their owner, thanks to outdated systems: manual spreadsheet records, ambiguous descriptions (“black umbrella”), and mountains of unsorted belongings at stations.
But in 2025, everything changed. With the Find platform, powered by artificial intelligence, Japan not only optimized the process: it tripled the recovery rate, reaching 30%.
How did you manage to turn something as frustrating as losing a wallet into an almost… easy process? The answer lies in a brilliant mix of accessible technology, common sense and a cultural obsession with solving real problems.
Lost property chaos in Japan
In Japan, where punctuality and order are the law, losing something on public transportation used to be a problem with no easy solution. Every year, more than 4 million lost items are reported in the Tokyo subway alone, from umbrellas to laptops and suitcases.
And although the Japanese are known for their honesty (90% of lost wallets are turned in to lost and found offices), the recovery system was incredibly inefficient. Before AI, the process depended on:
- Manual records: Employees wrote down vague descriptions on spreadsheets (“blue thermos”).
- Eternal searches: The owners had to look through shelves full of identical objects.
- Language barrier: Tourists faced forms only in Japanese.
The result: only 10% of the belongings returned to their owners. The rest ended up donated or destroyed after months in overcrowded warehouses. In 2023, the Tokyo Metro operator admitted that 70% of lost umbrellas were never claimed.
Losing something was not just a nuisance: it affected confidence in public transport and generated economic losses. But Japan, faithful to its philosophy of improving everyday life, found a solution where few were looking for it: in artificial intelligence.
Find: The technology behind the solution
The Find system revolutionized lost and found management in Japan by ingeniously combining artificial intelligence with existing infrastructure.
When someone drops off a lost item at one of the more than 2,000 participating stations and public spaces, the staff takes a photo that the system automatically analyzes.
AI doesn’t just record obvious features like color and shape, it detects unique details – a worn label, a particular scratch or even wear patterns – that turn each object into something identifiable among thousands.
For users, the process is surprisingly intuitive. Through popular apps like Line, they simply describe what they lost or upload a similar photo.
The magic happens when the system, powered by OpenAI’s GPT-4o, interprets even vague descriptions like “a black bag with a red detail near the closure” and matches them with its database.
In seconds, generate matches ranked by probability, overcoming the limitations of systems that relied on exact matches. The real innovation is in how Find integrates complex technologies into a flow that anyone can use.
Advanced image recognition and machine learning work together behind an interface as simple as sending a message. In addition, being available in four languages, it eliminates one of the biggest barriers for tourists.
Tangible results: From 10% to 30% success
Find’s impact on the recovery of lost items has been so overwhelming that the numbers speak for themselves. Before its implementation, companies like Keio Corporation were barely able to return 10% of lost belongings to their owners.
Today, that figure has tripled, reaching a remarkable 30% success rate. But beyond the percentage, what is truly revealing are the stories behind these statistics.
At Tokyo’s Yurikamome line, which once relied on manual records in spreadsheets, 78% of users who search for a lost item through Find now receive at least one potential match in less than an hour.
The most common items – such as umbrellas, wallets and headphones – have return rates exceeding 40%, a radical change for items that previously rarely returned to their owners.
The success of the system is measured not only in recoveries, but in time. What used to take days of paperwork is now resolved in hours: 65% of successful returns occur within the first 48 hours of the loss.
For foreign travelers, multilingual support has reduced cases of unclaimed items due to language barriers by 60%.
Why did Japan do it better?
Japan turned a universal problem into a success story thanks to a unique combination of cultural and technological factors.
Unlike other countries where lost objects often end up forgotten in drawers, here there is a culture of honesty and collective responsibility: 90% of the found objects are voluntarily delivered to the corresponding offices.
But what was really distinctive was that instead of creating a system from scratch, they took advantage of applications like Line and the network of train stations, giving immediate access to millions of users without having to learn new platforms.
Find is not the most advanced AI in the world, but it is the best adapted to a real problem. While others were developing complex chatbots, Japan solved something simple but frustrating: let your umbrella come back to you.
Lessons for the rest of the world
Japan has shown that the key to solving everyday problems with technology is not in complexity, but in accessibility and adaptation to the environment.
The implementation of artificial intelligence in the recovery of lost objects not only optimizes a bureaucratic process, but also reinforces trust in the system and improves the user experience.
Other countries can learn from this pragmatic approach: integrate technology into existing infrastructures rather than creating solutions from scratch, use intuitive tools for the public, and address real problems with precision.
The combination of technological efficiency and cultural values has turned a common problem into an effective solution.
Applying these principles could transform not only lost and found management, but many other aspects of urban life, from security to resource management, making the impossible seem simple.
This post is also available in: