Developing an embedded system solution for high-speed, high-capacity data logging for a size-constrained, low-power biomechanical telemetry system and investigating components for optimal performance.

摘要

The Purdue Neurotrauma Group (PNG) seeks to develop a biomechanical telemetry system capable of monitoring and storing athletes' head motions with the intention of identifying when a player may be at risk of neurophysiological damage, especially brain damage. A number of commercially-available systems exist with a similar goal; however, each of these systems discards information below an acceleration threshold. Research by PNG indicates that any acceleration may contribute to brain damage and that, because of this, an event-based model is insufficient for a proper understanding of an athlete's neurophysiological health. Continuous-time monitoring of head accelerations is therefore necessary. To facilitate the collection and storage of continuous telemetry data, a high-speed sensor system with a sufficiently large amount of memory storage is required. Additional requirements include low power consumption, low cost, and a small form factor. It has been concluded that a microSD card is the memory technology most capable of meeting these requirements, despite a number of drawbacks, most notably a relatively slow data write speed. An embedded solution requiring the use of large data buffers was developed to combat this drawback. Various microSD cards were tested to determine base read and write speeds and whether differences exist between card manufacturers, card sizes, or card speed ratings. It was found that the base performance was nearly identical in each test. Recommendations are made based upon the testing results, enabling production of operational prototypes for field evaluation.