Berenice Baker wrote Nebius for The Infinite Loop
Arriving passengers are greeted at Terminal B of the San Jose Mineta International Airport A humanoid robot named Jose. Mounted at a fixed base behind an information desk, Jose greets travelers and, switching to the language they speak, answers questions about flights, luggage and directions.
Users approach it hesitantly at first, then start testing it. Where they find baggage claim or switch languages in the middle of a conversation about a delayed flight. Some just want to know that Jose is a real autonomous robot and not controlled by a human behind the scenes.
Airports are heavily automated environments. Conveyor systems route baggage, software coordinates gates and exits, and automated ramp systems manage aircraft changes. Infinite loop by Nebiuss How the growing use of AI-powered humanoid robots working directly alongside humans is changing airport services.
San Jose Mineta (SJC) and Tokyo Haneda Airport (HND) is one of the busy hubs for robot trials in crowded terminals, runway aprons and passenger service areas. Some robots guide travelers in dozens of languages. Others are checked for baggage and ramp work.
The humanoid form factor isn’t just for innovation or branding. Human-shaped robots can work inside existing infrastructure with fewer modifications than many purpose-built automation systems.
Jose, the friendly face of Silicon Valley
But getting physically fit is the easy part. Passenger interactions make airports a difficult environment for public-facing AI systems. Information changes constantly, interactions occur under time pressure, and systems must function reliably despite noise and connectivity limitations.
This is why deployment is deliberately limited. Although Jose can stand and walk, IntBotThe company behind Jose, chose to tether the system during the pilot. According to the company, the decision was practical; In a crowded airport environment, loss of power or battery failure can create obvious safety issues.
While San Jose is hosting the match 2026 FIFA World CupThe airport’s four-month pilot is becoming an early test of how AI systems perform during periods of intense international passenger traffic.
The airport says Jose supports travelers in more than 50 languages, a capability that’s important in a linguistically diverse region like Silicon Valley. “What was really surprising to me was that I expected maybe 90% of the conversation to be in English,” Intobot product manager Hannah Wu said in an interview. “It actually turns out that 25% of interactions are in a language other than English.”
The system combines conversational AI with live flight and airport information, allowing passengers to naturally ask follow-up questions instead of navigating static menus or kiosks.
“Physical agent” orchestrating models across the edge and cloud
According to IntBot Chief Technology Officer Sharon Young, José distributes AI workloads across both the edge and cloud infrastructure depending on the task at hand. Instead of relying on a single unitary model, the system works as what Young describes as a “physical agent” by orchestrating multiple models and tools in real time.
Although Hose can access live flight and airport information in conversation with passengers, the system is relatively loosely integrated into airport infrastructure. IntBot said the system currently pulls flight status information and floor-plan data through APIs rather than acting as a deeply embedded backend system.
The difference reflects a larger challenge that physical AI deployments face. Airports are complex environments built around legacy systems, rapidly changing data, and human workflows that were not originally designed around autonomous machines.
Muki Patel, director of aviation at Mineta in San Jose, said the goal was not just to replace static information systems, but to test whether conversational AI could work effectively in a live terminal environment.
“Unlike a static kiosk or an app with pre-programmed responses, Jose can answer follow-up questions, change languages instantly and personalize interactions,” Patel said in an interview. “For first-time users, the interaction feels very similar to talking to a customer service agent.”
According to Hannah Wu, about three-quarters of interactions are socially driven rather than task-oriented, with passengers often approaching Joes out of curiosity before asking practical questions.
The company previously deployed robots in hotels before going to airports, in part because multilingual wayfinding and information services offered practical primary use cases for public-facing robotics.
“In a lab, there are so many edge cases that you can’t prepare for,” Wu says “In the real world, that’s how we make our products better.”
Japan Airlines Haneda Trial
Other airport robotics projects are focusing on labor-intensive tasks behind the scenes.
From the windows of Tokyo Haneda Airport’s departure lounge, passengers can see the familiar choreography of airport ground crews carrying unit load devices (ULDs), wedge-shaped containers used to load luggage, cargo and onto aircraft. But among the workers moving the containers onto the rotating ramps, one figure might look a little more metallic than the others.
Japan Airlines (JAL) and GMO are launching AI and Robotics Corp., which the companies describe as Japan’s first airport demonstration test Ground handling operations involve humanoid robots.
In contrast to the public-facing concierge role taken by Jose in San Jose, the Haneda trial focused on repetitive manual tasks already shaped around human crews. While initial deployments are focused on ULD transfer operations on airport ramps, future phases will potentially expand to baggage loading, cargo handling, cabin cleaning and even operations of ground support equipment.
The distinction is important. While humanoid robots are often touted as futuristic consumer technology, the strongest near-term case for deployment may be precisely the kind of routine work airports are increasingly struggling to staff.
“While airports appear to be highly automated and standardized, their back-end operations still rely heavily on human labor and face severe labor shortages,” said Tomohiro UchidaPresident of GMO AI & Robotics.
Japan’s aviation sector, like many of the country’s economies, faces a growing labor shortage driven by demographic changes and growing tourism demand. Ground handling work combines rigorous conditions with strict safety requirements and operational time pressures, which make automation attractive but difficult to implement using conventional industrial robotics.
According to JAL, the goal is not complete automation, but gradual workload reduction and productivity gains through systems designed to work alongside human crews. Using robots for physically demanding tasks “will inevitably reduce the burden on workers” and “provide significant benefits to employees,” Yoshiteru Suzuki, president of JAL Ground Services Company, said. told Kyodo News. Some duties, including security management, will still require human supervision, he added.
A JAL spokeswoman said the integration of humanoid robots and automated vehicles could eventually cut the need for workers in some container loading tasks by about half, contributing to the broader goal of improving productivity by 10% by 2030.
But scaling these systems beyond pilot projects will depend on proving they can operate reliably in highly constrained airport environments.
The airport doesn’t fit robots
The San Jose and Haneda projects reflect two different approaches to airport robotics. What they share is the challenge of deploying AI systems in environments built around human behavior and daily workflows.
This complexity is creating renewed interest in humanoid systems rather than purpose-built industrial machines.
The attraction of humanoid systems is not necessarily that they surpass conventional automation, but that they may be able to operate inside infrastructures already designed around human movement, tools, and procedures. In many cases, adapting robots to human environments may prove easier than redesigning airports around robots.
For now, most installations remain strictly off limits. Hose is stuck behind an information desk. The Haneda trial focused on limited operational tasks under controlled evaluation. The robot in the terminal is easy to photograph and easy to misread. It seems like the future has arrived. It’s actually an experiment in plugging the future into a building that was never designed for it
This is the story is produced by Infinite loop by Nebius and review and distribution Stacker.
–
Previously published at hub.stackernewswire
Good Men Project is pIndustry in a multi-platform ecosystem: flagship sites, Medium publications, substacks, social media. We publish experts, authors, therapists, coaches, activists and wellness professionals. AI tools look for stable, context-rich domains with a clear topical focus. GMP’s mission-driven content helps articles perform better in AI-driven discovery.
If you believe in the work we’re doing at The Good Men Project, please join us as a Premium Member today.
All premium members can watch The Good Men Project without any ads.
Need more information? A full list of benefits is here.
Photo credit: splash
