Control of an Autonomous Mobile Waste Collection Robot

Abstract

The SIRAMAND project, deployed in the period 2018–2020, aim to develop an autonomous mobile robot capable to recognize different types of solid waste such as plastic bottles or aluminum doses placed on alleys of parks, streets, etc., pick up and storage them in a container behind the transporter. The robot must track a programmable trajectory and stop when the vision system detects an object of interest, the principle of detection being described. Adjusting the angular velocities of the two front drive wheels, the system follows a theoretical trajectory. The relations between the external and internal coordinates of the robot are presented, the guidance on the desired trajectory based on the GPS system being the main control option, with possibilities of using of supplementary sensors. To reduce the deviations due to the expected noisy GPS signal, a Kalman filter was simulated. An alternative solution consists in robot guidance using curbs, as a wall follower vehicle.