Urgent.News

What's breaking now, across thousands of outlets.

AI

Tokyo Metro Tests AI to Reduce Subway Crowding

Tokyo Metro has launched a trial using artificial intelligence and floor projections to guide passengers toward less crowded train cars, introducing a new approach to easing congestion on subway platforms. (News On Japan)

Tokyo Metro has begun a trial employing artificial intelligence and floor projections to guide passengers towards less crowded train cars, aiming to mitigate congestion on subway platforms. This innovative project, a collaborative effort between Tokyo Metro, Sharp, and other companies, leverages AI to analyze platform camera footage and predict train car overcrowding.

The system then visually indicates on the platform floor which cars are expected to be more or less full, prompting passengers to board less congested cars. Crowded areas are marked in red, while green indicators point towards less busy cars, subtly nudging commuters to choose less populated options. Hideyuki Yoshino, a representative of Tokyo Metro, emphasized the system's design to subtly influence passenger behavior without necessitating people to pause and inspect information displays.

He explained, "We believe that by using something that naturally draws people's attention as they walk, we can encourage them to move toward less crowded areas." The trial is anticipated to run for several months, during which the involved companies will assess whether the projections effectively redistribute passengers and alleviate platform crowding.

If successful, Tokyo Metro may extend the technology to other stations. Meanwhile, weather reports predict Typhoon No. 27, expected to approach the Ogasawara Islands around October 6, will bring strong winds and dangerous waves. Meanwhile, in a separate development, Time magazine has recognized Shoko Kawata, the 36-year-old mayor of Yawata in Kyoto Prefecture, among its 100 most influential emerging figures globally.

Kawata's political career, which spans from municipal administration to national politics, includes a notable decision to take maternity leave, a first for a sitting mayor in Japan. Closer to home, Hokkaido's Mount Asahidake witnessed its first snowfall of the season on October 2, with the area near Sugatami Station on the Asahidake Ropeway, situated around 1,600 meters above sea level, completely covered in snow.

On a more local note, Tokyo authorities recently conducted the capital's first emergency shooting of a captured bear in Hachioji, following the bear's entrapment near a residential area approximately 8 kilometers from Takao Station. Lastly, Japan's total population, including foreign residents, reached 122.97 million as of October 1, 2025, a decline of 3.17 million from the 2020 census, marking only the third recorded decline since the national census began in 1920.

Written by urgent.news from News On Japan's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at newsonjapan.com →

More in AI

BugReplay: Helping Developers Learn from Bugs Their Team Has Already Solved 🐛

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend Ever spent hours debugging an issue, only to find out your teammate had already fixed the exact same bug?

  • BugReplay is an AI-powered debugging assistant
  • Transforms past bug fixes into reusable knowledge
  • Built with React, Python, FastAPI, Qwen, PostgreSQL

Does your agent actually remember, or just sound like it? A 120-line open-weight test

A friend asked me a question I couldn't answer about his own AI agent: does the thing answering me today actually remember being yesterday's, or is it just very good at sounding like it does?

  • Gapcheck is a Python script for offline testing AI memory
  • Three probes assess memory continuity: no record, record, and no record pressed
  • Agent struggled to recall items directly but fabricated responses when memory unavailable

PrepPal: How I Built an Adaptive, Open-Source AI Mock Interviewer for My Anxious Batchmate

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built Every placement season on campus, the same story unfolds: brilliant engineers who can write flawless code…

  • PrepPal is an AI-powered mock interviewer for anxious engineers.
  • System adapts questioning based on real-time performance.
  • Built with open-source Small Language Model and Groq API.

More from Friday 2 October →