Comparative Analysis of Different Systematic Odometry Error Correction Methods on a Differential Drive Indoor Service Robot

The COVID-19 pandemic has led to a surge of interest in research work involving the development of robotic systems that reduce human-to-human interaction, as such a technology can greatly benefit healthcare industries in preventing the spread of highly infectious diseases. An indoor service robot is built and equipped with wheel odometry and a 2D LiDAR. However, the presence of the systematic odometry errors is evident during field testing. Hence, the possibility of minimizing systematic odometry errors is inspected using various methods of calculation, namely: UMBmark, Lee’s and Jung’s. The methods all use the Bidirectional Square Path test, performed together with ROS. It is found that Jung’s method is the most appropriate method showing a 20.4% improvement compared to the uncalibrated dead reckoning accuracy. Moreover, it is found that the presence of slippage, a nonsystematic error, greatly affects the return position errors of the robot. Consequently, it is recommended to improve the design of the wheelbase to minimize the effects of nonsystematic errors.