Acoustic Assembly Verification
A robotic assembly application that evaluates the insertion of an electrical connector using the characteristic click sound. Microphone data is analysed to determine whether the connection was completed correctly.
Overview
This project investigates whether the acoustic signature of a connector 'click' can be used to verify correct assembly during a robotic insertion task.
Challenge
Visually confirming that an electrical connector has fully seated is difficult when the connector is small, partially occluded, or visually similar in both correct and incorrect states.
Solution
A microphone captures the insertion sound during robotic assembly, and a neural network classifies the recorded signal to determine whether the connection was completed correctly.
Implementation
- Robotic assembly setup using a Universal Robot
- Microphone placement and signal capture during insertion
- Signal processing of recorded audio
- Neural network classification of insertion outcomes
- Evaluation against manually labelled assembly trials
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.
Related publications
Related projects
Have a similar challenge?
Send us your technical requirements, drawing, or project idea and we will evaluate a practical approach together.