Implement to observe how agents move inside a tunnel full of opponents with minimum casualties. There are two strategies used. The first strategy called jump and go. When an agent sees an opponent in front of him, then the agent jumps to the opposite side of the opponent agent and moves forward. The second strategy is called wait and see. When an agent sees an opponent in front of him, the agent waits until the opponent passes out. Then move forward. After a few iterations, agents learn and select which behavior needs to be used to cross the tunnel from these two strategies.
Mesa - Multi agent simulator
pip install -r requirements.txt
python run.py
