Sign In

Bed sensor for sleep monitoring


We provide both algorithm methods and product prototyping for unobtrusive monitoring during sleep. Body movements, respiration, and heart motion can be detected unobtrusively with force-sensing foil located in a textile pillow or an under-the-bed mattress. VTT has improved this technology by developing a multi-channel force sensor array and related algorithms for extracting physiological signals. The prototype device has been tested for many different target groups. Sleep monitoring can be carried out at home for wellbeing applications or in hospital for detecting sleep disorders. The sensor array has also been placed in seats for measuring stress and relaxation during the daytime.

Sleep is a key component of healthy-living, but people may ignore it even during their normal life. In many cases, automatic monitoring-based sleep coaching could improve sleeping habits significantly, and thereby affect daytime efficiency or even prevent the risk of developing chronic diseases in the long term. Serious sleep disorders, such as insomnia or apnoea, have strong co-morbidities with cardiovascular diseases, diabetes and obesity. Detection and screening of sleep disorders already in their early stage would be important for the whole society.

Sleepiness and fatigue cause a considerable proportion of work- or traffic-related accidents, and recovery from stress is necessary to be able to work safely and efficiently. To help individuals in managing their daily routines, automatic monitoring methods using sensor arrays in seats have been developed in order to keep track of the periods of activity and relaxation.

Improving accuracy for contactless heart rate and respiration monitoring

Heart rate variability (HRV) is controlled by the autonomous nervous system, and HRV information can be used especially for detecting the level of parasympathetic activity. Based on this, VTT has developed models for both sleep staging and the detection of stress or relaxation during waking periods. However, good measurement resolution is essential, as the variation in the heart rate reduces with increasing age or e.g. due to chronic diseases. We have shown that the multichannel method improves accuracy of the heart beat interval measurement significantly when compared with a single channel sensor against the standard ECG electrode method.

Multiple movement sensors placed in different locations under the torso also give valuable information on the classification of the respiratory sleep disorders. In a typical apnoea event, the respiration effort may still be continuing while the standard clinical methods show the lack of respiratory airflow and reduced blood oxygen level. Finding these events of blockage in the respiratory airway from the respiratory movements only is complicated, but benefits from the multichannel movement detection. As a result, we have validated detection for the severity of the sleep apnoea with a VTT bed sensor in co-operation with sleep laboratories.


Related publications:

Kortelainen Juha M., Mendez M.O., Sleep staging based on signals acquired through bed sensor. IEEE Transactions on Information Technology in Biomedicine. Vol. 14 (2010) No: 3, 776–785.


Guerrero, G., Kortelainen, Juha M., et. al.: Detection of Sleep-Disordered Breathing with Pressure Bed Sensor, Conference proceedings, IEEE EMBC (2013), 3–7 July 2013, Osaka, Japan pp. 1342–1345.


Christoph Bruser, Juha M. Kortelainen, et. al.: Improvement of Force-sensor-based Heart Rate Estimation Using Multi-channel Data Fusion, IEEE Journal on Biomedical Health Informatics, issue 99, 2014.