Aarhus University Seal / Aarhus Universitets segl

Miscellaneous

Detect Mutual Influences

Unintended situations can emerge due to interactions of autonomous agents

Goal: 

  • Develop simple simulation environment
  • Use machine learningto detect influences
  • Use evolutionary algorithms to deal with external influences

Requirements:

  • Python or Java

Autonomous interaction & Team Formations

​Different agents can specialise on specific actions, form teams to perform tasks more efficiently  ​

Goal:

  • Learn own best actions and how to support others
  • Learn role in relation to other agents in the collective
  • Improve performance due to collaboration over individual operation

Requirements:

  • Programming

Neural Circuits

Generate simple neural circuit for a small robot based on sensed input

Goal:

  • Develop simple neural circuit based on the c.elegans
  • Connect neural circuit to robot
  • Robot performs actionsbased on input to circuit

Requirments:

  • Matlab Simulink
  • Programming

AI in games

Implement artificial intelligence (AI) in individual agents in popular computing games.

there are different games where we can do so, for example, implement how Ms PacMan should behave to evade all ghosts or how the individual ghosts should act in order catch Ms PacMan.

An alternative would be Star Craft. Here you can implement the behaviour of individual troops on the field and how they should behave in order to win over the players.

Goals:

  • Develop standard behaviour for agents
  • Use artificial intelligence and machine learning to improve behaviour

Requirements:

  • Programming (Python)