Research - 18.01.2019 - 00:00

Research project: warning system for diabetics on the road

Jointly with Inselspital Bern and ETH Zurich, ITEM-HSG is conducting research on a hypoglycemia warning system for diabetics on the road. The HEADWIND project will be funded by the SNSF with an amount of CHF 1.7m as from 2019.

18 January 2019. Success for a research team of ITEM headed by Prof. Elgar Fleisch and assistant professors Prof. Dr. Tobias Kowatsch and Prof. Dr. Felix Wortmann: Together with researchers from Inselspital Bern (Prof. Christoph Stettler, PD Markus Laimer and Dr. Thomas Züger) and ETH Zurich (Prof. Stefan Feuerriegel), they will receive a Sinergia Grant from 2019 onwards and thus research funding in the amount of CHF 1.7m from the Swiss National Science Foundation (SNSF) for an innovative research project at the interface between diabetes and modern automotive technology.

More road safety for diabetics

The overriding goal of the HEADWIND (Design and Evaluation of a Vehicle Hypoglycemia Warning System in Diabetes) project consists in a novel approach to an improvement of road safety for diabetes mellitus patients. Hypoglycemia can constitute a serious acute complication of a diabetes mellitus that is treated with insulin or certain other drugs. Hypoglycemia diminishes concentration, slows down perception and thought processes, and impairs numerous psychomotor functions. This is particularly critical on the road, where rapid decision-making sequences that integrate numerous factors are indispensable.

Data are recorded in real time

To reduce the increased risk of accidents of people with diabetes mellitus, the interdisciplinary and trans-university research team is going to break completely new ground and combine the immense opportunities of the rapidly developing automotive industry with innovative approaches from the field of artificial intelligence. The team of researchers intends to detect hypoglycemia directly from the data recorded by the vehicle on the road in real time.

Analysis with machine learning

Today, hundreds of driving parameters are recorded during a drive. Now these data are intended to be used and continuously analysed by means of so-called machine learning in order to recognise changes in the driver’s behaviour that would indicate hypoglycemia. In a first step, the researchers will conduct tests on a driving simulator, with hypoglycemia being induced in patients under medical supervision. In a next step, these tests will be moved to closed-off test routes, i.e. in real cars on roads. The great challenges of this project are not only data extraction and real-time processing with the application of complex mathematical algorithms, but also – and particularly – the controlled induction of hypoglycemia in a moving car, an endeavour which makes great demands on logistics and medicine and which is a "world premiere".

photo: peshkov –