Advancing tractor transmission performance through real-world data analysis

The global challenge of reducing greenhouse gas emissions and improving energy efficiency in agriculture has accelerated the need for data-driven methods in the design and assessment of tractor powertrains. This doctoral research, entitled Advancing Tractor Transmission Performance through Real-World Data Analysis, develops and validates a comprehensive methodological framework for analysing and optimising tractor transmissions using real-world operational data. The thesis is articulated in three main studies. The first introduces a non-invasive approach for estimating drawbar loads directly from CANBUS signals, enabling large-scale and long-term acquisition of traction data without additional sensors. The developed model—validated under field and transport operations—demonstrates good accuracy (errors within ±10%) and supports real-time analysis of traction efficiency and mission profiling. Building upon this foundation, the second study focuses on the optimisation of an input-coupled hydromechanical continuously variable transmission (IHMCVT). A parametric model was coupled with a constrained optimisation algorithm leveraging fleet-level data from over 300 tractors. Compared to traditional design approaches, the data-driven optimisation increased transmission efficiency by +0.67% and output power by 1.9% during representative duty cycles, proving the value of embedding usage statistics into drivetrain design. Finally, the third study proposes the Fleet-based Real-Load Output Cycle (FRLOC), a synthetic yet statistically robust duty cycle that reproduces the operational variability of agricultural tractors. Derived from CANBUS data, FRLOC covers up 92% of the real traction area and maintains dynamic continuity across field and transport phases. It provides a realistic benchmark for assessing tractor performance and validating simulation tools. Overall, the thesis bridges the gap between theoretical modelling and real-world tractor operation. By integrating data analytics into drivetrain design and testing, it contributes to the development of more efficient, adaptive, and sustainable agricultural machinery. The outcomes support the transition toward data-informed engineering practices, offering tangible benefits in terms of fuel savings, emission reduction, and system optimisation.