This assignment focuses on developing an intelligent system for task and motion planning in a 2D environment. The objective is to enable a robot (R) to collect assignment reports from four regions and deliver them to a submission desk while minimizing motion costs. The environment consists of a 6m x 6m space, with each region assigned a waypoint. Randomly sampled waypoints are interconnected to create a roadmap. The robot's path is determined by the Euclidean distance between waypoints. It means that if the robot reaches the corresponding waypoint location of the region that region will be considered as being visited. The entire project can be found in the attached Zip.file
The first step in running the project successfully is to install the popf-tif planner (which can be found in this repository).
If you use the docker ubuntu image, take in mind that the planner is based on the Ubuntu-22.04 Docker Official Image, which lacks a graphical interface. To change the planning files, use the following command to mount a shared directory between your host machine and the docker image:
docker run -dit -v path/in_your/host/folder:path/in/docker_container/folder --name your_name hypothe/ai4ro2_2
The docker image given by Dr.Antony Thomas , is the simplest method to execute the project. This image already includes all of the necessary setup to execute the project. Simply extract the visits_domain and visits_module folders included with this repository and replace them in the '/root/ai4ro2' directory.
The visits_domain folder contains a PDDL domain as well as issue files, whereas the visits_module folder contains the files required to execute the external planner.
The poptf repository : https://github.com/popftif/popf-tif
The docker image given by Dr.Antony Thomas : https://hub.docker.com/r/hypothe/ai4ro2_2
The "random" library from c++ : https://cplusplus.com/reference/random/
Aicha Manar ABBAD - s5565902
Ines HAOUALA - s5483776
Roumaissa BENKREDDA - s5434673
Karim TRIKI - s5528602