Structures and
Autonomy Laboratory

Built infrastructure, such as bridges and roads, is fundamental to daily life, yet it faces various risks and challenges throughout its service life. Ensuring its safety and functionality requires both technical expertise and practical engineering judgment. Each time a fundamental technological breakthrough emerges, the field has sought to incorporate it under severe practical constraints, including challenging field conditions and limited resources, to maximize its impact on infrastructure safety and performance. Today, rapid advances in AI, big data, and autonomous systems are creating new opportunities. Our aim is to propose and demonstrate pathways that connect these fundamental breakthroughs to people's daily lives through improved management of built infrastructure.

Research themes for incoming students

We welcome individuals interested in addressing challenges facing critical infrastructure through innovations in both technical foundations and practical deployment. Research themes include, but are not limited to:

• 2D and 3D computer vision for infrastructure condition assessment: image-based structural component and damage detection, point cloud-based automated geometric modeling, and damage quantification and mapping for structural digital twins.
• Autonomous decision-making for bridge inspection agents: mobile robot navigation, embodied question answering, and vision-language-action modeling for bridge inspection.
• Platforms and benchmarks for autonomous infrastructure inspection systems: automated generation of synthetic environments and datasets, unsupervised domain adaptation, and sim-to-real transfer.
• Data-assimilative decision-making for enhanced infrastructure resilience: assimilation of computer vision and other sensing modalities, and integration of information at the network level.

Please feel free to contact Dr. Yasutaka Narazaki if you have questions or are interested in joining the group.