![[2601.02085] Vision-Based Early Fault Diagnosis and Self-Recovery for Strawberry Harvesting Robots](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2601.02085] Vision-Based Early Fault Diagnosis and Self-Recovery for Strawberry Harvesting Robots
Summary
This article presents a framework for early fault diagnosis and self-recovery in strawberry harvesting robots, leveraging vision-based technology to enhance efficiency and stability during operations.
Why It Matters
As agriculture increasingly adopts automation, ensuring the reliability of robotic systems is crucial. This research addresses common faults in strawberry harvesting robots, potentially improving yield and reducing labor costs, which is vital for the agricultural sector's sustainability and productivity.
Key Takeaways
- Introduces SRR-Net, a multi-task perception model for strawberry harvesting.
- Achieves high accuracy in strawberry detection, segmentation, and ripeness estimation.
- Implements corrective control strategies to mitigate common harvesting faults.
- Utilizes real-time visual feedback for improved grasp classification.
- Demonstrates competitive inference speed while maintaining perception accuracy.
Computer Science > Robotics arXiv:2601.02085 (cs) [Submitted on 5 Jan 2026 (v1), last revised 15 Feb 2026 (this version, v2)] Title:Vision-Based Early Fault Diagnosis and Self-Recovery for Strawberry Harvesting Robots Authors:Meili Sun, Chunjiang Zhao, Lichao Yang, Hao Liu, Shimin Hu, Ya Xiong View a PDF of the paper titled Vision-Based Early Fault Diagnosis and Self-Recovery for Strawberry Harvesting Robots, by Meili Sun and 5 other authors View PDF HTML (experimental) Abstract:Strawberry harvesting robots faced persistent challenges such as low integration of visual perception, fruit-gripper misalignment, empty grasping/misgrasp, and strawberry slippage from the gripper due to insufficient gripping force, all of which compromised harvesting stability and efficiency in orchard environments. To overcome these issues, this paper proposed a visual fault diagnosis and self-recovery framework that integrated multi-task perception with corrective control strategies. At the core of this framework was SRR-Net, an end-to-end multi-task perception model that simultaneously performed strawberry detection, segmentation, and ripeness estimation, thereby unifying visual perception with fault this http URL on this integrated perception, a relative error compensation method based on the simultaneous target-gripper detection was designed to address positional misalignment, correcting deviations when error exceeded the tolerance this http URL mitigate empty grasping/misgrasp and fruit-slip...