面向复杂水下环境的智能机器人研究与工程实践
Research and engineering for intelligent underwater robotic systems

• Novel sensing — sensing principles, calibration and underwater data acquisition
• Task equipment — robotic platforms, manipulators and operation equipment
• Perception & control — robust perception, control and autonomous inspection
• Navigation — localization, planning, search and multi-vehicle coordination
• Deep-learning foundation — reusable architectures across the four directions

Our stable research groups are intentionally small: one graduate lead with one or two highly motivated members. We separate long-term research routes from time-bounded projects, and we turn reliable results into seven classes of reusable technical assets.

• Define the question and acceptance evidence before implementation
• Keep code, configuration, data, model and hardware versions traceable
• Record failed experiments and root causes as carefully as successful results
• Require independent review before an asset is promoted for reuse
• Build documentation that the next researcher can actually follow

Public research repositories and reusable assets will be released after internal review. Organization members can use the member view for internal navigation.

_{Maintained by the HIT Underwater Robot research team.}