Related Experiment Video
Updated: Mar 11, 2026
07:15
Tactile Vibrating Toolkit and Driving Simulation Platform for Driving-Related Research
Published on: December 18, 2020
5.1K
Usability Evaluation of Warning Cues for Remotely Supervised Autonomous Agricultural Machines
Anita C Ezeagba1, Sebastian Lorenz2, Cheryl M Glazebrook3
1Department of Biosystems Engineering, University of Manitoba, Winnipeg, Manitoba, Canada.
Journal of Agricultural Safety and Health
|March 9, 2026
Summary
User experience significantly impacts autonomous agricultural machine (AAM) warning systems. Both visual-auditory and visual-tactile modalities are usable, but effectiveness depends on the supervisor's background, influencing interface design needs.
Area of Science:
- Human-Computer Interaction
- Agricultural Technology
- Human Factors Engineering
Background:
- Increasing prevalence of autonomous agricultural machines (AAMs) necessitates effective remote supervision systems.
- Warning systems are crucial for ensuring safety and efficiency in AAM operations.
- Evaluating multimodal warning systems is essential for optimizing human-machine interfaces (HMIs).
Purpose of the Study:
- To evaluate the usability and cue effectiveness of visual-auditory (VA) and visual-tactile (VT) bimodal warning modalities for AAM supervision.
- To assess how user background, specifically farming experience, influences the perception and effectiveness of warning cues.
- To determine the impact of warning modalities on situation awareness at perception, comprehension, and projection levels.
Main Methods:
- A simulated remote supervision task was conducted with 30 participants (5 with farming experience, 25 without).
- Usability was measured using the Computer System Usability Questionnaire (CSUQ).
- Cue effectiveness was assessed via objective performance measures including response time, accuracy, and comprehension/projection times.
Main Results:
- Both VA and VT modalities demonstrated high usability, with farming participants favoring VT and non-farming participants favoring VA.
- Non-farming participants showed significantly faster response times to VA warnings.
- Farming participants exhibited lower projection accuracy with VA cues, though overall comprehension and projection times did not significantly differ by modality.
Conclusions:
- User background, particularly farming experience, is a critical factor in determining the effectiveness of AAM warning systems.
- The study highlights the need for adaptive and user-informed HMI designs that cater to diverse user experiences.
- Both VA and VT modalities are viable, but optimal choice depends on the specific user group and task requirements.
