Real-time work progress estimation based on GIS remote monitoring system for agricultural robot vehicles
Ricardo Ospina, Noboru Noguchi
Abstract
The remote operation of multiple agricultural robots improves the efficiency of farming operations, but monitoring dispersed robots over long distances is challenging. The estimation of work progress is difficult due to the variability of task completion times influenced by terrain conditions. Adjusting multiple robots’ velocities within set time limits under unpredictable field conditions lacks sufficient research and practical validation. To address these challenges, this research proposes a remote monitoring system based on a Geographic Information System (GIS), with a focus on estimating work progress and effectively controlling two agricultural robot vehicles. The main innovation of this research lies in the system’s capability to estimate the total time required for tasks and calculate work progress based on real-time data from the robot vehicles. Furthermore, it adjusts the robots’ operational velocity automatically based on the work progress, optimizing task efficiency. The GIS-based remote monitoring system has been tested on real agricultural robots, demonstrating its capability to handle multiple vehicles simultaneously while maintaining low latency. Simulation results over the internet indicate that the system’s delay is around 100 ms, ensuring safe operation when a maximum of 50 machines are used simultaneously. Experimental results using real robot vehicles validate that the overall system’s delay is around 2 s. Thanks to the automatic velocity set point adjustment of the system, the robot vehicles were able to complete their work plan closer to the initially estimated time. Without the automatic velocity set point adjustment, the vehicles were, on average, 26 s ahead or 33 s behind schedule. However, with the automatic velocity set point adjustment in place, the vehicles’ time deviations were reduced to just 13 s early or 4 s late, resulting in a time error reduction of 48–87 %.