IMAC-AgriVLN: Can Agricultural Vision-and-Language Navigation Agents be Aware of Instruction Mistakes?
2026-06-01 • Robotics
Robotics
AI summaryⓘ
The authors address the problem that agricultural robots often follow natural language instructions assuming they are always correct, which is not realistic because instructions can have mistakes. They created a new benchmark, A2A-MI, that introduces common instruction errors to test robots more realistically. They found that current robot navigation systems struggle with these errors, showing a big drop in performance. To fix this, the authors designed a module called IMAC that helps the robot detect and correct mistakes in instructions during navigation, improving its performance significantly. Their work highlights the importance of enabling robots to question and adjust instructions when navigating.
Agricultural robotsVision-and-Language Navigation (VLN)Natural language instructionInstruction mistakesA2A-MI benchmarkInstruction mistake awarenessIMAC moduleNavigation performanceData annotationRobot instruction correction
Authors
Xiaobei Zhao, Xingqi Lyu, Xin Chen, Xiang Li
Abstract
Agricultural robots are serving as powerful assistants across a wide range of agricultural tasks, nevertheless, still heavily relying on manual operations or railway systems for movement. The AgriVLN method and the A2A benchmark pioneeringly extended Vision-and-Language Navigation (VLN) to the agricultural domain, enabling a robot to navigate to a target position following a natural language instruction. However, almost all the prior methods adopt an ideal assumption that the given instructions themselves are correct, which does not align with the realistic scenarios, because anybody may say an instruction with mistakes. To bridge this gap, we propose the A2A-MI benchmark, in which we build a semi-automatic data annotator to insert three mistake classifications into each original instruction in a more diversified and efficient way. We test several state-of-the-art agricultural VLN agents on it and observe a sufficient drop with -57% on SR and -9% on NE, from which we suggest that an agricultural VLN agent tends to assume that the given instruction is correct, so does not have the awareness to doubt it when the scenes it sees do not align with the instruction it receives. To build the awareness on instruction mistake, we propose the IMAC module analyzing the instruction and the current front-facing image, to judge whether the instruction has mistakes and attempt to correct it when needed. We integrate IMAC into the baseline model, and observe a noteworthy improvement, sufficiently narrowing the gap to the performance on instructions without mistakes. Project: https://github.com/AlexTraveling/IMAC-AgriVLN.