Assignments developed for the Intelligent And Mobile Robotics class. The assignments can be found on archive.
The robotic agents were developed using the CiberRato simulation environment. The easiest way to test this code is to add this repository as a subomdule of a CiberRatoTools repository. First, change directory into the ciberRatoTools repository, and then run the following command:
git submodule add git@github.com:diogopjesus/rmi-assignments.git agent
Then, go into the agent directory and build the agents:
./build.sh
Finally, run the agent (X is a value between 1 and 4, corresponding to the challenge):
./run.sh -cX [-h host | -r robname | -p pos | -f outfile] # X between 1 and 4
To test the scores of the generated best paths and perceived maps, run the following:
./test-run.sh -cX [-f outfile] # X between 1 and 4
To automate runs, you can apply the automateRuns_CiberRatoTools.patch (available on patches) to the ciberRatoTools repository, and then run the loop.sh script (do not forget to re-build the simulator).
./loop.sh -cX # X between 1 and 4
3 different robotic agents, each one that aim to solve one of the following problems:
- Localization,
- Mapping,
- Planning.
Programmed in Python.
Final Grade: 17.1/20
Develop a single agent that solves the three problems presented in assignment 1 with only a line sensor, a noisy compass and the movement model.
To approximate the values of the compass a Kalman filter was implemented.
Programmed in C++.
Final Grade: 16.7/20