Leakage-Aware Robotic Grasping
Development of a robotic grasp optimisation system for vacuum grippers. Pressure sensor data, robot information, and geometric data are used to evaluate grasp quality and improve the reliability of robotic handling processes.
Overview
This project investigates how vacuum-gripper grasp quality can be estimated from pressure sensor readings combined with robot and geometric information, rather than from vision alone.
Challenge
Vacuum grippers can fail silently: a seal may look correct visually while still leaking air, leading to dropped or misplaced objects during automated handling.
Solution
Pressure sensor data is combined with robot pose and geometric information in a machine learning pipeline that estimates grasp quality and flags likely leakage before an object is moved.
Implementation
- Instrumentation of a Universal Robots cell with vacuum gripping and pressure sensing
- Collection of grasp trials across varied object geometries
- 3D vision-based geometric feature extraction
- Machine learning model development in Python
- Experimental evaluation of grasp-quality predictions
Technologies
Gallery
Technical diagram
Results
- Placeholder result — replace with a verified project outcome.
Limitations
- Placeholder limitation — replace with a verified constraint of the current approach.
Next steps
- Placeholder next step — replace with a verified planned extension of this work.
Related publications
Have a similar challenge?
Send us your technical requirements, drawing, or project idea and we will evaluate a practical approach together.