Development and Implementation of Laboratory Test Methods for the Evaluation of Wearable Head Impact Sensors

With a rise in wearable sensor technology and the desire to investigate head impacts in previously unstudied groups, wearable head impact sensors have reached nation-wide popularity for their promising benefits to consumers and researchers. However, there are risks in relying on such technology before proper validation of its performance has been completed. Preliminary tests have found that current sensors vary widely in performance. The objective of this work was to develop and implement a test method for evaluation of wearable sensors in an ideal laboratory environment. A custom pendulum was used to impact a NOCSAE headform mounted on a Hybrid III neck. Sensors were tested under helmeted and unhelmeted conditions, according to their prescribed use. The headform was impacted at four locations, each at four impact energies ranging from 25 g to 100 g. Peak and time series headform kinematics output by each sensor were compared to accelerometers and angular rate sensors inside the headform. Average and standard deviations of peak sensor error and normalized RMS error were evaluated at each test condition to describe sensor performance. Requirements were set in the slope and coefficient of determination from linear regressions constrained through the origin to describe adequate sensor performance under ideal conditions. Sensors that met the requirement in at least one kinematic variable will be further evaluated in more realistic on-field and cadaver tests. The combination of all testing phases will be used to provide an overall sensor evaluation for both researchers and consumers.