First Workshop on
Robotic Computing for Industrial Inspection

15:30 - 15:40

Welcome and Opening Remarks

15:40 - 16:00

Development of Autonomous Mobile Robot for Social Infrastructure Inspection

Because of population decline and aging, the lack of personnel for maintenance and inspection of social infrastructure is becoming a worldwide problem. In this presentation, we introduce the development of an autonomous mobile robot to automate indoor/outdoor patrol in social infrastructures, such as plants, power facilities, construction sites, and so on. This presentation provides a brief overview of our developments, such as off-road capable drive mechanism, 3D measurement and reconstruction of the surrounding environments, autonomous navigation and meter monitoring. Furthermore, we explain key technologies to achieve stable navigation even in unknown ground environments, namely online detection of low-level step and route planning considering step entry direction.

16:00 - 16:20

Making the world a safer place with robotics

Robotics has emerged as a pivotal solution for inspection and maintenance tasks in confined spaces, where human intervention can be perilous. This presentation focuses on the Float Arm, an innovative robotic system successfully deployed at the Fukushima power plant. By enabling remote operation, the Float Arm enhances safety for operators, allowing them to oversee critical tasks without entering hazardous environments. This technology not only protects human lives but also significantly reduces maintenance costs through increased efficiency and precision. Additionally, the integration of robotics in such operations supports broader digital transformation efforts, enabling real-time data collection and analysis that drive informed decision-making. We will explore the impact of the Float Arm's deployment, highlighting its contributions to safety, cost-effectiveness, and the future of automated maintenance in challenging settings.

16:20 - 16:50

Paper Pitch Session

Authors will give brief, engaging overviews of their research papers, highlighting key points and significance.

16:50 - 17:10

Autonomous Drone Navigation in Urban Non-GPS and GPS Hybrid Environments Using CityGML and BIM

In urban infrastructure development and disaster inspections, there is a demand for safe and efficient technologies. However, challenges exist in utilizing drones in urban areas with numerous safety constraints. Drone operations in urban environments face issues such as positioning errors due to GPS multipath effects occurring between buildings, making reliance on GPS alone insufficient and necessitating redundancy (duplication or triplication) in positioning functions. Additionally, there are challenges like the need to create large-scale pre-mapped point cloud data of the operational area for LiDAR-based self-position estimation. In this study, we created an integrated map by combining outdoor and indoor building data using CityGML and BIM, and adopted a hybrid method for self-position estimation that combines GPS, LiDAR, Visual-Inertial Odometry (VIO), and an Inertial Measurement Unit (IMU). This approach allowed us to eliminate the pre-mapping point cloud data creation process and directly utilize high-precision urban models for navigation. In self-position estimation, accurate positioning became possible by supplementing with LiDAR-based localization in areas where GPS is unstable. Furthermore, for vertical movements along building walls—which tend to be unstable when relying solely on LiDAR localization—the introduction of VIO improved robustness. Through experiments conducted in actual urban environments, we confirmed that it is possible to eliminate the effort required for pre-map creation and achieve stable, continuous autonomous drone navigation both indoors and outdoors. This research contributes to enhancing the efficiency and safety of drone operations in urban areas.

17:10 - 17:30

Aerial Robotic Inspection of GPS-denied and confined spaces

In recent years, teleoperated and semi-autonomous solutions have been developed to inspect critical infrastructures, such as ballast tanks and cargo hulls. Thanks to their 3D workspace, Uncrewed Aerial Vehicles (UAVs) are particularly suitable for deployment in confined areas to protect human surveyors and workers from the high risks of the most dangerous part of the inspection. This talk addresses the main technical challenges towards autonomous aerial robotic inspectors, in the context of two EU H2020 projects: AutoASSESS and Aviator. In particular, the following will be discussed: the mechanical design of compact multi-rotor UAVs, the integration of NDT sensors, the development of portable modules for controllability, and navigation with onboard sensors.

17:30 - 17:50

Coffee break

17:50 - 18:00

Closing Remarks

18:00 - 18:30

Poster Session and Discussion
[in-person only session]

Authors will present their posters in person at the workshop location, providing an opportunity for direct interaction and in-depth discussion of their research.