Workload Assessment of Tractor Operations with Ergonomic Transducers and Machine Learning Techniques

Area of Science:

• Ergonomics and Human Factors
• Occupational Health and Safety
• Agricultural Engineering

Background:

• Dynamic muscular workload assessments for tractor operators are under-researched, impacting operator efficiency and safety.
• Repeated clutch and brake operations in tractors impose significant physiological and ergonomic loads.

Purpose of the Study:

• To assess and model the dynamic muscle load, physiological variations, and discomfort experienced by tractor operators.
• To evaluate the effectiveness of wearable ergonomic transducers and data-run techniques for workload assessment.
• To explore the use of machine learning for predicting operator discomfort and optimizing machinery design.

Main Methods:

• Nineteen tractor operators performed tasks on three tractor types across varied speeds and surfaces.
• Electromyography (EMG) and custom foot transducers measured muscle load and actuation forces.
• Heart rate (HR), oxygen consumption rates (OCR), and energy expenditure rates (EER) were recorded.
• Operator-reported overall discomfort ratings (ODR) were collected, and machine learning models were applied.

Main Results:

• EMG revealed high muscle strain (e.g., gastrocnemius right at 43% MVC), exceeding recommended levels.
• Clutch and brake forces were significantly influenced by operating speed, tractor type, and surface (p < 0.05).
• Energy expenditure rates (EER) indicated moderate-heavy to heavy physical workload (9-24 kJ/min).
• Machine learning models (KNN, RFC, SVM) predicted ODR with high accuracy (87-97%), with RFC being most effective.

Conclusions:

• Existing tractor clutch and brake systems may require design refinement to reduce operator strain.
• Energy expenditure and actuation forces correlate well with EMG signals, suggesting simpler workload assessment methods.
• Machine learning offers a robust approach to objectively assess and predict operator discomfort, informing future ergonomic designs for agricultural machinery.