A study to develop a smartwatch-based blood glucose monitoring and warning system for people with diabetes

The aim of this study is to investigate whether smartwatch data can be used to detect hyperglycemia and hypoglycemia as early as possible. The latest smartwatch models can record a variety of medically relevant data. This includes, for example, heart rate and heart rate variability, movement and diurnal patterns. This data will be analyzed using machine learning in order to warn the user in real time of hyperglycemia and hypoglycemia, which can lead to serious medical complications. This solution is a non-invasive (i.e. without injuring the body) alternative to common methods of detecting excursions of blood sugar. This is particularly interesting for people with diabetes who cannot or do not want to wear a continuous glucose sensor, yet want to receive early warnings of hyperglycemia and hypoglycemia.

Study results

Financial Support

  • Innosuisse:  Swiss Innovation Agency 


  • ETH Zurich: Prof. Dr. Elgar Fleisch, Dr. med. Thomas Züger, Martin Maritsch, Simon Föll
  • University of St. Gallen: Prof. Felix Wortmann

Principal Investigator