Autonomous Driving AI Laboratory
Department of Information Systems, Faculty of Engineering, Saitama Institute of Technology

Autonomous Driving AI Laboratory
(Watabe Lab)

Saitama Institute of Technology, Fukaya, Saitama, Japan

Principal Investigator: Prof. Daishi Watabe

Professor, Department of Information Systems, Faculty of Engineering, and Graduate School of Engineering
Vice President, Saitama Institute of Technology
Director, Autonomous Driving Technology Development Center

Degree
Dr. Sci. (Tohoku University)
Email
dw@sit.ac.jp
researchmap
researchmap.jp/daishiwatabe
ORCID
0000-0003-2731-2226
Scholar
Google Scholar

About the lab

The lab conducts research and education on AI-based image recognition for autonomous driving and biometrics. Since late 2018 we have developed autonomous buses for public roads on the open-source Autoware stack, working across localization, perception, planning and control, fail-safe design, and the steering and pedal actuators themselves.

The lab is the core of the university's Autonomous Driving Technology Development Center (established 2019), which Prof. Watabe directs. Our buses carry passengers in regular service, and the driving logs they record feed back into our AI research.

Members

The group of about 40 people is led by Prof. Watabe and works with the Autonomous Driving Technology Development Center on vehicle development, field operation, and data analysis.

PositionNumberRole
Principal Investigator (Professor)1Leads the lab and the Autonomous Driving Technology Development Center
Doctoral students3Doctoral research on autonomous driving
Research engineers (PhD, lab alumni)2Autonomous bus development at the Development Center
Research assistant (part-time)1Research support
Safety driver1On-board operation of the autonomous buses
Undergraduate students (4th year)14Graduation research
Undergraduate students (3rd year)17Pre-graduation research and seminars

Research

  • Public transit

    Autonomous buses in regular service

    We have taken part in developing five autonomous buses with support from national, prefectural and municipal governments and a foundation. Since April 2025 our Isuzu Erga Mio bus has served the Fukaya City community bus route, covering its full 37 km loop. A large autonomous school bus has run between the campus and the nearest station since 2022.

  • Machine learning

    Learning-based planning from real driving logs

    The buses have recorded about 8,000 runs (about 66 TB) of LiDAR, 4D radar, camera, GNSS/IMU and vehicle CAN data. We use these logs to train and tune diffusion-model trajectory planners for Japanese bus operation, and compare them with vision-language-action models on the same scenes, using the ABCI supercomputer.

  • Safety

    Localization and fail-safe design

    Switching between LiDAR- and GNSS-based localization along urban routes, detection of sensor faults, and redundant emergency-stop paths that remain effective when software fails.

  • Maritime

    World's first self-driving amphibious bus

    In 2022 we developed, with partners, unmanned driving and navigation technology for an amphibious bus that drives on land and sails on Lake Yamba-Agatsuma, Gunma.

  • Biometrics

    Ear recognition robust to pose

    Ear biometrics that stay reliable under out-of-plane rotation, from a single enrolled image, supported by three JSPS KAKENHI grants (2010–2017).

Selected publications

  1. Enhancing Safe Driving of Autonomous Buses in Obstructed-View Conditions Using Distributed Monocular Cameras. IECON 2024, 50th Annual Conference of the IEEE Industrial Electronics Society. doi:10.1109/IECON55916.2024.10905580
  2. Improvement of Self-Localization Sensor Transition Based on Autonomous Driving. IECON 2023, 49th Annual Conference of the IEEE Industrial Electronics Society. doi:10.1109/iecon51785.2023.10311843
  3. World's first self-driving amphibious bus. International Robotics & Automation Journal. doi:10.15406/iratj.2023.09.00254
  4. ACDR: Autonomous-car drive recorder. Journal of Robotics and Mechatronics, 2020. doi:10.20965/jrm.2020.p0634
  5. Joystick Car Drive System and its Application to Self-driving Microbus. IECON 2020, 46th Annual Conference of the IEEE Industrial Electronics Society. doi:10.1109/IECON43393.2020.9254364
  6. Use of Generated Ear Images by GAN to Biometrics Based on Deep Learning. IEICE Transactions A, 2020. doi:10.14923/transfunj.2020BAL0002

Full list (77 papers): researchmap

Professional service

  • Member, Working Group on Safety Standards and Inspection, Study Group on Autonomous Ships, Maritime Bureau, Ministry of Land, Infrastructure, Transport and Tourism (2024–2025)
  • Member, ISO/IEC JTC 1/SC 37 Biometrics (2016–2025); Secretary, WG4 and WG6 (2018–2025)
  • Member, IEICE Technical Committee on Biometrics (2014–)
  • Editorial Planning Board, Institute of Image Information and Television Engineers (2015–)
  • Program Chair, International Conference on Biometrics and Kansei Engineering (2019)

Teaching

Information and Coding Theory; Statistical Processing I; Basic Information Engineering Laboratory; Applied Programming Languages; Advanced Media Engineering (graduate).

Contact

Prof. Daishi Watabe
Department of Information Systems, Faculty of Engineering, Saitama Institute of Technology
1690 Fusaiji, Fukaya, Saitama 369-0293, Japan
Email: dw@sit.ac.jp · Tel: +81-48-585-2940 (Autonomous Driving Technology Development Center)