▼Job Description
Design and build scalable data collection systems operating across hundreds of humanoid and mobile robots
Develop low-latency, semi-autonomous, and bilateral teleoperation systems that maximize controllability while minimizing operator workload
Build automated pipelines for large-scale model deployment and rigorous evaluation in both simulation and real-world robotic platforms
Collaborate closely with the Vision-Language-Action (VLA) team to integrate and deploy state-of-the-art models onto physical robots
Engineer robust robotic software and hardware systems for reliable real-world operation
Diagnose, debug, and optimize performance across perception, control, and system integration layers
Translate research prototypes into scalable, production-grade engineering systems in collaboration with cross-functional teams

■Benefits
Few initiatives globally match the scale of this project in collecting large-scale robotic data and developing foundation models for embodied AI. As part of one of Japan’s leading national programs, the project is supported by a 20.5 billion yen investment from NEDO.

This role will be central to the project’s success. You will have significant ownership and technical autonomy, with the opportunity to shape core systems that define the future of large-scale robotics and embodied AI.

Engineers are strongly encouraged to grow their careers through this initiative—for example, by leading system deployments, contributing to open-source robotics software, and driving engineering excellence across large-scale robotic platforms.

・A New Challenge for the AI Robotics Association (AIRoA)

The AI Robotics Association (AIRoA) is launching an ambitious initiative to collect one million hours of humanoid robot operation data from hundreds of robots and use it to train the world’s most powerful Vision-Language-Action (VLA) models.

What makes AIRoA unique is not only the unprecedented scale of our real-world data and humanoid platforms, but also our commitment to making everything open and accessible. We are building a shared “robot data ecosystem” where datasets, trained models, and benchmarks are available for anyone to use.

・What this means for researchers

For researchers, this means the opportunity to:
Tackle fundamental challenges in robotics and AI, including multimodal learning, manipulation with rich tactile feedback, sim-to-real transfer, and large-scale benchmarking
Access state-of-the-art infrastructure: hundreds of humanoid robots, GPU clusters, high-fidelity simulators, and a global-scale evaluation pipeline
Collaborate with leading experts from academia and industry, and publish that will shape the next decade of robotics
Contribute to redefining the future of embodied AI—while all results are shared openly with the world

Flextime system

■Required Qualifications
M.S. or PhD in Robotics, Mechanical Engineering, Electrical Engineering, Computer Science, or a related field, with 3+ years of relevant industry experience (or equivalent hands-on experience)
Extensive hands-on experience with complex, high-DOF robotic systems, including bimanual mobile manipulators and humanoid robots operating in real-world environments.
Strong experience controlling real robotic systems using ROS/ROS2, including full-stack system integration and deployment
Solid understanding of teleoperation systems, AR/VR interfaces, and leader–follower control architectures
Strong programming skills in C++ and Python, with the ability to write production-grade code and perform system-level integration
Demonstrated ability to work effectively both independently and within multidisciplinary engineering teams
Experience developing, validating, and deploying reliable robotic systems in production or field environments

■Preferred Qualifications
Publications in top-tier robotics conferences or journals (e.g., TRO, IJRR, RSS, CoRL, ICRA)
Experience taking complex robotic systems from prototype to production
Practical experience with tactile sensing and contact-rich manipulation
Experience contributing to large-scale robot data collection or fleet management systems
Familiarity with robotic simulators such as Isaac Sim or MuJoCo for training and evaluation
Experience integrating Sim2Real transfer methods or Vision–Language–Action (VLA) models onto real robotic systems
Experience with CI/CD pipelines and system validation methodologies for robotics applications
Experience building full-stack web applications (frontend and backend) for robotics data infrastructure, evaluation platforms, or fleet management systems

Fluent (TOEIC 865-)

Daily Conversation Level

Social Insurance

Five-Day Workweek
Summer Holidays
Winter Holidays
Congratulatory or Condolence Leave
Child-care Leave
Paid Holidays

Full-time employment