# Human vs. Teleoperated Robots in Vineyard Management: A Simulation-Based Analysis of Travel Speed, Routing, and Task Performance

arXiv: https://arxiv.org/abs/2507.04167 · HTML: https://mlsanity.com/paper/2507.04167/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Jul 2025.

## Built on by (most cited first, 1 of 1)

- [Taxonomy and Trends in Reinforcement Learning for Robotics and Control Systems: A Structured Review](https://mlsanity.com/paper/2510.21758/) (2025)

## Abstract (arXiv)

Rising labor costs and narrow treatment windows have made teleoperated robots a proposed tool for vineyard scouting and treatment, but simulation-based comparisons of human and robotic performance are sensitive to modeling choices that can favor one agent before any trial runs: whether both can record a target's location, whether both are routed efficiently, and whether travel speeds are assessed consistently. This paper simulates a human operator against a teleoperated robot on two vineyard tasks, disease detection-and-treatment and yield-zone survey, holding both agents to matched location-marking and routing standards, using a routing strategy proven optimal for this topology, and modeling teleoperation-specific effects. Once information access and routing are held equal, task-time outcomes are governed largely by relative sustained travel speed: at the speeds modeled (1.4 m/s human, 0.9 m/s UGV), the human completes both tasks 43-48 percent faster, a gap persisting across a simulated eight-cycle season, while at speed parity the agents differ by only 9-12 percent, and every other parameter shifts the result by under 4 points. These results indicate the comparison is primarily a question of platform speed, and that the case for teleoperated robots should rest on labor availability, endurance, consistency, or hazard avoidance rather than completion time.