ASABE Technical Library - Abstract
Member and Access Notice
Kalman Filter-Based Sensor Fusion for Site-Specific Mechanical Weeding System
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: 2026 ASABE Annual International Meeting 2601011.(doi:10.13031/aim.202601011)Authors: Shafi Md. Istiak, James Y. Kim, Cristhian Perdigon, Sulaymon Eshkabilov
Keywords: Site-specific mechanical weeding, Kalman filter, individually controlled sweep, sensor fusion, weeding robot. Introduction
Abstract The site-specific mechanical weeding system (SSMW) is a three-row sweep-based autonomous weeding robot developed for targeted between-row weed control in row crop production systems. The system was designed to perform autonomous individually controlled tillage operations using structural strain-based feedback control. A strain-based load cell was integrated into the sweep shank to measure the strain generated during soil engagement, while a laser distance sensor was used to monitor sweep position and assist in identifying sweep-soil interactions. Laboratory and field experiments revealed nonlinearity, drift, and uncertainty in the strain-based load cell signal. To improve the reliability and stability of the feedback signal, a Kalman filter-based sensor fusion approach was developed by integrating the strain-based load cell and laser distance sensor signals. The laser sensor provided a stable positional reference, while the strain sensor retained the dynamic soil engagement information. The proposed Kalman filter-based sensor fusion method was implemented and evaluated using MATLAB/Simulink with field-acquired sensor data to assess the feasibility of reliable feedback estimation for an individually controlled sweep tillage operation. The results demonstrated an improved estimation of soil-sweep engagement.
(Download PDF) (Export to EndNotes)