Enhancing Logistics Automation: Integrating Capacitive Proximity and Tactile Sensors for Trolley Pose and Center of Mass Estimation

Warehouse capacity usage and damage rate are some of the most essential metrics in logistics. In this paper, we propose a concept using capacitive sensors to reduce passage width and ensure robust material handling. We present preliminary results for trolley pose estimation in a drive-in task and a proof-of-concept for trolley Center of Mass estimation using a capacitive sensor system. We propose a markerless and infrastructure-agnostic method for relative trolley pose estimation and updating the mobile robot’s localization. The sensors are mounted on an autonomous mobile robot and measure capacitive fluctuations in the electromagnetic field. Two capacitive proximity sensor units with 16 electrodes are mounted on top of the robot. The proposed method is based on Principal Component Analysis and allows for I-shaped profile pose estimation based on proximity feedback. To evaluate the proposed method, a robotic arm was introduced to capture the profiles above the sensor array and provide ground truth. The sensors were then mounted on a mobile robot and evaluated in a drive-in task with comparable results to the state-of-the-art vision-based approach. Under the sensor array, four additional tactile sensing modules are installed to estimate the object’s center of mass using the multi-point weighing method, achieving repeatability of about ±50g.