Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Real-time Artificial Intelligence for a robot billiard player
This project is part of a new IfA project to develop a fully automated billiard playing robot, and is in particular concerned with its strategic AI. It will use a previous successful project as a baseline to implement a real-time AI for the robot. To start in September 2019 or earlier.
This project is part of a new IfA project to develop a fully automated billiard playing robot. Strategy in billiard games is complex, because a) the outcome of a particular shot is uncertain, b) the range of legal shots is large and continuous-valued, and c) it is sometimes difficult for a player to decide where he/she would want the balls to be after the shot has been taken. This project will break the decision problem down using methods from dynamic programming (DP) and tree search algorithms. Both methods characterise future rewards (in this case points scored relative to the opponent) as a function of the system state, in order to help the player plan ahead.
This project is part of a new IfA project to develop a fully automated billiard playing robot. Strategy in billiard games is complex, because a) the outcome of a particular shot is uncertain, b) the range of legal shots is large and continuous-valued, and c) it is sometimes difficult for a player to decide where he/she would want the balls to be after the shot has been taken. This project will break the decision problem down using methods from dynamic programming (DP) and tree search algorithms. Both methods characterise future rewards (in this case points scored relative to the opponent) as a function of the system state, in order to help the player plan ahead.
The steps of this project will be as follows:
1) Review the report and associated AI code from the previous Master's Thesis on this topic. This consists of a mixture of direct dynamic programming and Monte Carlo tree search algorithms.
2) Create a data pipeline that links the state estimation (ball locations) from the robot's existing vision system to the AI unit.
3) Develop a decision-making algorithm that can operate within the 1-2 minute timeframe permitted for a real-world billiard shot.
4) Test the algorithm by passing the shot commands to the robot arm.
5) Evaluate the effectiveness of the algorithm in selected strategic scenarios.
The steps of this project will be as follows:
1) Review the report and associated AI code from the previous Master's Thesis on this topic. This consists of a mixture of direct dynamic programming and Monte Carlo tree search algorithms. 2) Create a data pipeline that links the state estimation (ball locations) from the robot's existing vision system to the AI unit. 3) Develop a decision-making algorithm that can operate within the 1-2 minute timeframe permitted for a real-world billiard shot. 4) Test the algorithm by passing the shot commands to the robot arm. 5) Evaluate the effectiveness of the algorithm in selected strategic scenarios.
Joe Warrington (warrington@control.ee.ethz.ch), Nikos Kariotoglou (karioto@control.ee.ethz.ch)
Joe Warrington (warrington@control.ee.ethz.ch), Nikos Kariotoglou (karioto@control.ee.ethz.ch)