About 425 million adults around the world were living with diabetes in 2017. A relevant condition called Hypoglycemia is characterized by a dangerous low level of blood sugar that could be fatal to diabetic patients. Continuous glucose monitoring systems (CGMS) are the most popular commercially available technology for detecting diabetic hypoglycemia. However, CGMSs are invasive, costly, and not user-centric thereby not sustainable for diabetes management. This paper documents our initial efforts in designing an inexpensive, non-invasive, wearable physiological tremor sensory system to detect the onset of hypoglycemic events of diabetic patients. The design cycle briefly presented here includes: 1) determination of system (technology and user) requirements, 2) development of the tremor detection prototype, and 3) testing and validation of the system in non-clinical and clinical settings using human factors, data analytics, and biomedical sciences techniques and approaches.