2021/08/03 by Ekaterina Karmanova, Valerii Serpiva, Karmanova, Ekaterina +7
Computer Science · Social Sciences · #Digital Games and Media #FOS: Computer and information sciences #Human-Computer Interaction (cs.HC) #Reinforcement Learning in Robotics #Robotics (cs.RO)
paper · pdf · doi:10.48550/arxiv.2108.01593
openalex publication_date 2021/08/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Reinforcement learning (RL) methods have been actively applied in the field\nof robotics, allowing the system itself to find a solution for a task otherwise\nrequiring a complex decision-making algorithm. In this paper, we present a\nnovel RL-based Tic-tac-toe scenario, i.e. SwarmPlay, where each playing\ncomponent is presented by an individual drone that has its own mobility and\nswarm intelligence to win against a human player. Thus, the combination of\nchallenging swarm strategy and human-drone collaboration aims to make the games\nwith machines tangible and interactive. Although some research on AI for board\ngames already exists, e.g., chess, the SwarmPlay technology has the potential\nto offer much more engagement and interaction with the user as it proposes a\nmulti-agent swarm instead of a single interactive robot. We explore user's\nevaluation of RL-based swarm behavior in comparison with the game theory-based\nbehavior. The preliminary user study revealed that participants were highly\nengaged in the game with drones (70% put a maximum score on the Likert scale)\nand found it less artificial compared to the regular computer-based systems\n(80%). The affection of the user's game perception from its outcome was\nanalyzed and put under discussion. User study revealed that SwarmPlay has the\npotential to be implemented in a wider range of games, significantly improving\nhuman-drone interactivity.\n