In safety-relevant autonomous systems, obtaining a result is only the beginning. Students must test their assumptions, identify failure modes, explain uncertainty and communicate evidence for reliability. Across our subjects we organise learning as a progression from principles to implementation, validation and reflection: students work with reproducible code, simulation, real sensor data and open-ended engineering tasks rather than treating AI, navigation or control as purely abstract topics.

Lectures, demonstrations, code review, project milestones and assessment follow the same cycle we use in research: specify assumptions, implement a method, evaluate it against evidence, diagnose its limitations and defend the result. Early low-stakes exercises reveal misconceptions in coordinate frames, sensor models, stability and optimisation before students commit them to a larger project; annotated examples and guided labs support students from different backgrounds, while extension tasks stretch experienced programmers. This supports inclusion without lowering standards.

We build a connected pathway from secondary school to research and industry. Through hands-on robotics outreach and the PolyU Junior Researcher Mentoring Programme, secondary-school students try AI, coding, drones and autonomous vehicles. At PolyU, courses lead into capstone and URIS projects, and we support the PolyU AI & Robotics Club, where students from engineering, construction, fashion and design work in mechanical, electronics, perception and AI groups and progress from training to Robocon and RoboMaster teams, internships and research.

Students have won a Third Prize in the Hong Kong University Student Innovation and Entrepreneurship Competition, a Merit Award for Best URIS Research Project, the OneRobotics track championship at the 2026 Hong Kong Physical AI Hackathon, and competed in the China Universities Aircraft Design Competition.