The project main goal is to map buildings with the help of autonomous remote control cars, equipped with high resolution camera and ROS based computer. This package deals with the AI module of the project, which upon receiving a 2D array, returns the best coordinate to travel to.
- Modular classes enable the implementations if various strategies of finding the most optimal coordinate.
- Flexible and adaptable implementation of the board.
This package is a part of software project directed by Ben Gurion University. The project combines both Software Engineering department and Electrical Engineering department.
Knowing the surface structure is crucial and critical routine of many professionals such as firefighters and security forces, receiving information about the operating surface in the most efficient and accurate manner is often a matter of life or death.
The goal of our project is to map structures using vehicles and creating a computer model that describes, as accurate as possible ,the operating surface. When being activated the vehicle, which is equipped with various cameras and sensors, will be located at the entrance of the structure and will navigating autonomously in order to create the most accurate model of the structure. By doing so, this model can help saving human life when entering into dangerous areas.
Deployment is very simple, just extract the compressed file into the designated computer.
car's computer should have all of the files except for client_app.py under the same folder, where the client_app.py file should be located at the client-side computer.
As mentioned before, this system is divided into car-side and a client side.
Running of the driver-side components is done with the execution of init.sh which takes care of running all of the built-in
classes.
all configuration that can be altered are inside the file named conf and can be edited in form of a Json (later will be translated to python dictionary using ast library.
The configuration file is being used by both car-side and client-side programs
In order to run the client-side program
python client_app.py
In order to stop the whole process, run the script stop.sh which will close all of the driver-side terminals.
The whole scanning process is being controlled by the client-side terminal (will be the base for app development).
After a successful connection to the car-side, type start to start the scanning process after which direction will appear on the screen.
Press ok if following the direction is done successfully, press fail if the steps were not fulfilled completely.
for stopping the scanning process press stop
Currently there are 3 main computing engines that compute the most optimal coordinate to visit by different factors.
engine V2: Will prefer the cell marked as unmapped that is located in front of the car's direction and has the greatest distance
engine V3: Very similar to engine V2 but will prefer the cell that also has the greatest number of unmapped neighbors
This whole project is written by the following members:
- Gal Porat
- Osher Damari
- Tomer Belzer