Assembly line fatigue monitoring

September - October 2019

A HAX Startup developing light passive exoskeleton supporting the arms and shoulders wanted to give their customer more insights on worker fatigue and task monitoring on overhead assembly lines
I developed a simple wearable they could strap on the exoskeleton to measure:

fatigue from repetitive movements or maintained posture
long term exposure to vibrations from tool use
exposure to noise

Step 1: Data gathering prototypes

To get an idea of what sort of data we would encounter using wearables for activity monitoring, I built a few sensor breakout boards to use with a Raspberry Pi Zero W, with which I could do basic preliminary data gathering and analysis using python. I used the data gathered with these prototypes to narrow down the final sensor selection (body coupled 9 axes IMU and microphone) and choose the general signal processing approach to use in the next iteration.

IMU breakout board is fastened on an elastic band and was tested in different locations of the arm. Ideally, it would be mounted before the elbow, on the exoskeleton itself.

Step 2: Embedded wearable sensor node

After the sensor approach was validated I focused on making a lower power, smaller module to be deployed onsite for a longer study. The sensor node is based on an NRF52832 and can log/and stream data over BLE.