Assessing and Ensuring Sensor Reliability in Agricultural Robotic Systems

The Navion Project focuses on improving the autonomous navigation of robots in the agricultural sector. Hanze University of Applied Sciences is collaborating with various partners, each with their own challenges, such as indoor/outdoor navigation, irregular or flat terrain, connectivity, and the ability to accurately determine position.

This project examines methods for evaluating sensor reliability and performance during operation. Students will develop and test algorithms for detecting common sensor issues such as blur in camera images, data loss in LiDAR scans, or noise in radar responses. The project also explores sensor fusion checks to verify cross-consistency and the potential for real-time detection of corrupted or obstructed sensor signals.

Research questions
• How can we automatically monitor and assess sensor quality and reliability in autonomous agricultural robots operating under variable conditions?
• How is the reliability of sensor data affected by the operation of the robot in greenhouses when compared to outdoors?
• LIDAR: How many datapoints are useful? Where is the threshold?

Deliverables
1. A framework for sensor health monitoring, improving the safety and robustness of autonomous navigation systems.
2. A paper/report documenting methods, findings, and recommendations, including performance analysis.
3. GitHub repository with all the code and instructions to use it.