Page 49 - THEi Student Applied Research Presentations 2025

CLIMBSMART: IMPLEMENTATION OF AI IN
INDOOR ROCK CLIMBING

Poster
CHEUNG Kai Sun Kasun
BSc (Hons) in Information and Communications Technology
Department of Digital Innovation and Technology

RESEARCH BACKGROUND OBJECTIVES

Rock climbing has been gaining traction in cities like Hong Kong, especially • Make indoor climbing safer and
with more indoor climbing gyms popping up. While it’s a fun and rewarding smarter
sport, it comes with its share of risks—like falls, overuse injuries, and technical • Help climbers avoid injury and
mistakes. improve their skills
• Promote rock climbing in Hong
METHODOLOGY Kong

Firstly, the user captures a video input using a smartphone, which is then
preprocessed. Next, an AI-based hold detection system uses YOLOv8 to locate
holds. The system then moves to route mapping and performance metrics
to estimate their difficulty, analyze path optimality and assess performance
smoothness. Lastly, it generates feedback and reports for user to improve.

FINDINGS

ClimbSmart demonstrates strong potential to enhance indoor climbing with
AI-powered route detection and performance feedback. Early tests show high
accuracy and fast processing speeds, proving its feasibility for real-world use.
With refinements to the UI, data diversity, and real-time tracking, ClimbSmart is
on track to become a valuable tool for safer, smarter climbing.
ABOUT THE INVESTIGATOR

My interests are rock climbing,
coding, experimenting with generative
AI, singing and gaming. My ultimate
career goal is to launch my own
product and start a company. But for
now, I aim to find a company working
with AI that I can learn from.
FYP Supervisor: Mr. LAU Ho Chuen,
Tommy

Student Applied Research Presentations 2025 Student Applied Research Presentations 2025 38