Humanoid robots start sorting luggage in Tokyo airport test amid labor shortage

· Source: AI - Ars Technica · Field: Transportation & Mobility — Robotics & Autonomous Systems, Logistics & Freight Transportation, Public Transportation & Urban Mobility · Depth: Fundamental Awareness, quick

Summary

Japan Airlines is initiating a demonstration project at Tokyo's Haneda Airport, running from May 2026 to 2028, to test humanoid robots for tasks such as baggage handling and cargo loading, aiming to mitigate a significant human labor shortage. This experiment, a collaboration between JAL Ground Service and GMO AI & Robotics Corporation, will deploy Unitree Robotics' G1 and UBTECH Robotics' Walker E robots, leveraging advanced AI models to adapt to complex airport environments. While humanoid robots offer potential cost-effective assistance, with models like the Unitree G1 costing as low as \$13,500, their effectiveness in unpredictable settings and safety considerations when working alongside humans remain critical challenges. The initiative addresses a severe labor deficit, exemplified by Narita Airport's inability to service over 30 percent of requested flights in late 2023 due to staff shortages.

Key takeaway

Japan Airlines is piloting Unitree G1 and UBTECH Walker E humanoid robots at Tokyo's Haneda Airport from May 2026 to 2028 to address a severe labor shortage in baggage and cargo handling. This trial aims to determine if AI-powered humanoids can adapt to unpredictable airport environments, despite current limitations in demonstrated effectiveness and significant safety considerations. The initiative provides critical insights for aviation, logistics, and robotics professionals on deploying general-purpose automation in complex, human-centric operational settings.

Topics

Best for: Executive, Investor, Entrepreneur, Robotics Engineer, Operations Professional, Tech Journalist

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.