The project involves research on court between a real player practising a variety of shots and technical aspects of their swing, a motion tracking sensor attached to the squash racket to monitor swings and speeds of hitting the ball, and a robot receiving this data and interpreting what coaching feedback to provide.
“We want to discover if robotics can complement and support the activities of a human coach, bridging the gap between in person coaching sessions when an individual sportsperson is conducting solo practice,” said Martin Ross, a PhD student at Heriot-Watt University. “Robots have been used in other sports, but we believe this is a world first for squash coaching. We are looking forward to sharing results on whether a robot can improve different aspects of a player’s game and increase their motivation for solo practice. Additionally, we are observing the human response to a robot coach, assessing whether machine-led coaching strategies and words of encouragement are accurately timed and positively received.”
The project is being led by scientists from the National Robotarium, which is hosted by Heriot-Watt University, and is being delivered in collaboration with industry partner, RacketWare, using its motion tracking sensor and interfacing technology to translate a player's data to the robot.
Twelve coaching strategies have been developed from observing one-to-one training sessions between real coaches and players. Based on the sensor’s data, the coaching strategy programming will enable the robot to replicate the live coaching experience by giving instructions about what the player should practice next, using hints and tips and providing positive reinforcement.
Dr Paul Mellor, founding director of Racketware, said: “A coaching session is typically £40 per session with an estimated 10-20 coaching sessions required to make progress. It’s expensive, especially if you’re committed to improving your game.
“That’s why we’re excited about this project. It has the capacity to push the frontiers of how motion sensor technology can be applied to racket sports and explore the area of robotic coaching with motion sensor data to develop products, like ours, that make squash coaching accessible to more people.”
“We know that solo practice offers great benefits to squash players so if the use of a robot coach can motivate more players to try it then it will be a great thing for the sport in Scotland,” added Paul Bell from Scottish Squash, the national governing body for the sport in Scotland.
Testing started in July at Oriam, Scotland's national performance centre for sport, based at Heriot-Watt University's Riccarton campus in Edinburgh. The project will be completed in March 2022.