Individual Squash Training is More Effective and Social with a Humanoid Robotic Coach*
Martin K. Ross (Heriot-Watt University)
Frank Broz (TU Delft - Interactive Intelligence)
Lynne Baillie (Heriot-Watt University)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
With the aim of providing extra motivation to adhere to repetitive, individual sports training, this paper presents an autonomous robotic squash coach capable of high-level personalisation. The system was evaluated in person with 16 participants each conducting three 15-minute solo practice sessions. We compared a baseline, non-coaching robotic condition to two conditions in which the robot executed one of 12 different coaching policies, each of which was based on human coaching data. In one of the coaching conditions, the policy was selected based on categories for personalisation and in the other it was selected randomly among policies. The coaching policy conditions were found to be more enjoyable, more socially competent, and perceived as a more effective coach than the baseline.