Development of a Smart Lawnmower with a Mapping System Utilizing GPS and LiDAR Technology

Abstract

This paper reports on the design, implementation, and field evaluation of an autonomous lawnmower that combines GPS, IMU, and LiDAR sensing within a ROS2-based control architecture to support real-time mapping, path planning, and obstacle avoidance. The hardware platform and control architecture are outlined, including a PID motor controller and an Extended Kalman Filter used for multi-sensor fusion. Two navigation configurations; Nav2 paired with the SLAM Toolbox and Nav2 paired with GPS are evaluated and compared in terms of lawn-mowing accuracy, obstacle-avoidance performance, and the ability to pass through narrow gaps. Over six outdoor trials per configuration, the SLAM Toolbox attained a mean navigation accuracy of 88.6% (SD = 7.9%), significantly higher than the 62.1% (SD = 11.1%) mean accuracy achieved by the GPS-based configuration (Welch's t-test, t = 4.76, p = 0.001); nevertheless, because GPS-based navigation does not depend on fixed structural landmarks, it proved more robust for open-field operation. The SLAM Toolbox also recorded a perfect 100% (6/6) obstacle-avoidance success rate against 80% (8/10) for GPS, and was able to pass through gaps as narrow as 40 cm. Environmental testing further showed that LiDAR mapping quality was affected by ambient light, with noise levels lower during night-time trials. Together, these results point to a trade-off between the mapping precision offered by SLAM and the landmark-independence offered by GPS, and they motivate future work on RTK-GPS integration, full-coverage path planning, and camera-based sensor fusion to strengthen robustness for open-field lawn maintenance.