Mars Rover

Mars Rover Project (autonomous navigation and arm movement)

Autonomous Navigation

ROS navigation stack is used while performing autonomous navigation. The rover uses Raspberry Pi cameras for science experiment monitoring, manipulator control feed and AR tag detection. The rover can detect multiple AR tags together. AR tag detection works quite well from 4 to 5 meters and in low light conditions. The rover uses an extended Kalman filter localization node and a navsat transform node to fuse IMU and GPS data together and determine rover’s position with respect to its local co-ordinate frame. Rover’s real time position can be seen overlayed on Satellite map. Ultrasonic sensor is used to obtain 180-degree scan data for determining obstacles and perform SLAM gmapping. Rover’s software uses DWA and navfn for path planning. The rover can track AR tags and determine the tags pose. Shortly after the SAR, a stereo camera will be used implement 3D point cloud based mapping and visual odometry, which will greatly improve the rover’s autonomous performance. 

Autonomous arm control

Inverse kinematics control is used for the 6 DOF robotic arm. Our manipulator can turn a knob, open a latch, lift a weight, toggle a switch, open a drawer and insert a container into the drawer, pressing mechanical Keyboard precisely with the end effector head and operate a joystick, pull a rope, tighten a screw as shown in the SAR video. Firstly, the on-body sensors will allow us to align the rover to the servicing station for equipment servicing tasks. The gyro sensor, camera feedback, and AI-based operations running in the background will help the manipulator align with better accuracy. Inverse kinematics will track down the motions and position of the gripper.