UCLA

Optical spectroscopy technology has evolved significantly, from its use in research to detectnew materials to its use in the medical field to diagnose diseases. Applications of this technology have been used in wearables that provide insights to promote health and wellness. Rockley Photonics has prototyped a novel wearable via a miniaturized laser system, and the characterization of the laser output power is important for algorithm development. The functional data of interest includes repeated measurements across multiple laser systems operating at various current and temperature ranges. The ability to develop a repeatable and reliable test system is crucial, and thus this paper aims to apply statistical methods to propose a framework in developing models for power prediction. Prototype laser systems (N=5) were tested using a standard procedure and laser output power was characterized to understand significant test factors. Preliminary findings suggest that operator and repeated measurement factors are not significant to the output power and that models can be developed on an individual laser channel bases using an ordinary least squares (OLS) model with third order current effects. The application of using the same base model for individual lasers on each tested device is important for continued development of algorithms to be used to detect various biomarkers in the health sensing space.