EV3Rover is a data acquisition module for an any-time and any-space, sparse-sensing SLAM system operating on low-cost robotics platforms. It facilitates environment exploration and constructs scans of the environment using an ultrasonic sensor.
You need a LEGO Mindstorms EV3 robotics kit. Using the components found in this kit you'll need to consturct a non-holonomic, differential-drive vehicle. To aid with hardware assembly, the below LeoCAD images are provided:
After constructing the hardware vehicle you'll need to install the EV3DEV Operating System onto your EV3 brick. Instructions on how to do this can be found on the EV3DEV webiste.
Once you have installed the EV3DEV OS you will need to clone and build the EV3DEV-LANG-CPP library.
git clone https://github.com/ddemidov/ev3dev-lang-cpp.git
cd ev3dev-lang-cpp
mkdir build
cd build
cmake .. -DEV3DEV_PLATFORM=EV3
make -j4
make install
cd ../..After you have installed the above library, clone and build this repository:
git clone https://github.com/sg2295/EV3Rover.git
cd ev3rover
mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=MinSizeRel
make -j4
./ev3rover # Run the executable (if you're on the EV3 device)It is recommended to do the above in a Docker image of the EV3DEV OS, instead of the EV3 hardware, to speed up compilation. After building this repository, copy the executable over to the EV3 brick via scp and you're ready to go!
N.B. the design of EV3Rover is inherently modular and allows for the trivial swapping of both the motion and observation modules. If you choose to use a different vehicle or sensor configuration, you'll need to update the respective module(s) accordingly.
EV3Rover uses googletest for unit testing. Note that due to unsatisfied CMake requirements, we use the (now outdated) 1.11.0 release of gtest.
In case of any problems, please feel free to get in touch with me using the email found on my github profile, or by opening a ticket on the EV3Rover repository!
I created EV3Rover as part of my BSc Computer Science dissertation at the University of Bath. EV3Rover formed the basis of my research into the development of an any-time and any-space, sparse-sensing SLAM platform. It was used for environment exploration and data acquisition.