data compression for Smart Sock

Sensor data pre-processing & compression for BLE transmission


Problem Statement: 

This assignment is part of the project “Expressure” which measures the pressure on human feet using “smart socks”, for medical research. The “smart sock” prototype is the outcome of our work-package, mainly which consists of an Arduino-Nano and a matrix pressure sensor. It needs to collect 70 sensor-readings per-foot data in real-time at a frequency of 100HZ. These data then needs to be transmitted through BLE to servers (laptop/phone). The prototype and corresponding software are almost finished, but one of the problems is that the BLE transmission speed becomes a bottleneck, which is slower than the sampling rate.

Tasks: 

Students are expected to increase the transmission speed through either data compression or pre-processing algorithms. A research on sensor data compression and pre-processing techniques are firstly suggested. Subsequently, students can implement their algorithms on the device for evaluation.

Several factors can be considered when addressing this problem:\

  1. Part of human feet may not touch or touch softly on the ground when standing or walking, so that not all 70 sensor give information, many readings could just be zero.
  2. When people are standing or sitting still, the readings are quite static.

Students need only to program and test the algorithm on Arduino, others such as the sensor/hardware/server are already prepared.


Work: 

40% Theory, 30% Programming, 30% Writing


Contact:  

Wei Wang (W.Wang-1@utwente.nl)