Exploration of an unknown environment is a fundamental problem in the field of
autonomous mobile robotics. It deals with the task of examining the environment with
sensors while creating a map with its collected data. Conventionally, humans map the
environment beforehand in order to use theresult in a vehicle for subsequent navigation.
However, exploration has the potential to remove humans from the loop of generating a
map of an unknown environment. This has many applications such as robot deployment in
areas where pre-mapping is not possible, space robotics, etc..
The main goal of this project is to design a robot control software for a Robosoft Kompai robot in the simulated factory environment MRDS with a hybrid deliberative/reactive architecture,allowing the robot to navigate the environment while constructing a map of it.
- Launch a simulation environment in MRDS
- cd to the directory where the source files are contained
python3 main_test.py
The exploration area was limited in order to avoid small areas, this is because the resolution of the world was not enough to traverse it without bumping into objects; doubling the resolution resulted in deliberation times in the order of a minute. The chosen resolution lead to a good trade-off.