Publication

On the limit: Estimating round baler torque for grass harvesting scenarios

Crotty, Tomás
Warren, John
Shanahan, Conor
Jones, Edward
Glavin, Martin
Byrne, Dallan
Citation
Crotty, Tomás, Warren, John, Shanahan, Conor, Jones, Edward, Glavin, Martin, & Byrne, Dallan. (2025). On the limit: Estimating round baler torque for grass harvesting scenarios. Computers and Electronics in Agriculture, 232, 110056. doi:https://doi.org/10.1016/j.compag.2025.110056
Abstract
This paper introduces a novel, cost-effective approach to measure loading torque on an agricultural baler implement. Round balers gather grass fodder and form cylindrical bales, which are subsequently used for feeding livestock. As fodder accumulates within the bale chamber, the torque on the drive shaft increases, necessitating an increase in input power. The operator must manage the loading levels to prevent excessive strain on implement components and avoid a breakdown. State-of-the-art torque measurement techniques are difficult to integrate into harvesting implements due to their significant cost and mechanical footprint. Consequently, baling implements lack loading measurement capabilities, forcing operators to gauge power draw without quantitative feedback. This paper proposes a sensor solution to approximate torque on a round baler, significantly reducing costs when compared to existing methods. The approach correlates chain stress throughout bale formation by monitoring deformation of the chamber drive chains. Measurements were acquired from a fixed-chamber baler-wrapper with ground-truth torque benchmark recordings. Experiments were carried out using artificial loading on a static machine and then within the field, with a mobile implement forming grass silage bales. A prototype sensor was installed on a grass harvesting contractor’s baler-wrapper to examine performance across a single season, across 180 sites in the West of Ireland. Results from each scenario highlight a linear relationship between sensor and ground-truth torque measurements. The method achieved a 7% mean peak torque estimation error, allowing operators to monitor mechanical loading limits during bale formation and to increase efficiency in grassland agricultural
Publisher
Elsevier
Publisher DOI
Rights
Attribution 4.0 International