The accurate solution for human activity tracking

The growing trend of the quantified-self has generated a large offer of products on the market. Despite the high competition, sector-leading customers have been attracted by our large IP portfolio, by the precision of our extensive algorithm library, as well as by our extended know-how in the development of wearable low-power devices. The rapidity in adapting our software solutions to new sensor locations (e.g., wrist, thorax, head, foot), to new applications, or to different external hardware platforms has made CSEM an indispensable partner for a successful product.

This human activity tracking technology is part of a larger effort to advance Digital Health, offering real-time health insights through wearable solutions.

CSEM’s comprehensive algorithm library includes 

Feature - Description

• Activity classification - Automatic classification of user's activity among [sleep, rest, walk, run, bike, swim].
• Step count - Steps executed during walk and run activity classes.
• Total sleep time - Amount of sleep time in a sleep episode.
• Sleep stage scoring - Classification of sleep (actigraphy-based) among [wake, light-sleep, deep-sleep].
• Cadence - Instantaneous cadence estimation during walk, run, bike, and swim activities.
• Traveled distance - Traveled distance estimation during walking, running, and swimming.
• Speed - Speed estimation during walking, running, and swimming (swimming-pool).
• Workout - Session duration of vigorous physical exercise or training.
• Energy expenditure - Amount of energy burned by the user.
• Swim stroke count - Number of swim strokes per lap. Stroke means the number of hand entries, left and right combined.
• Swim lap count - Number of swam laps.
• Swim style classification - Automatic classification of swim style: butterfly/crawl, backstroke, and breast stroke.
• Swim efficiency - Indirect measurement of swim efficiency called SWOLF.
• Fall detection - Automatic detection of fall events.

Security

These features are computed in real time within the wearable device leaving a minimal memory and battery footprint. The development listed above also includes preprocessing expertise of analog sensor signals to ensure robust extract features, namely the enhancement of optical signals for cardiac activity monitoring. These CSEM’s software solutions are regularly benchmarked against commercial devices in clinical-like conditions and are continuously being improved in order to keep ahead of the market.

More information

• Delgado-Gonzalo, A. Lemkaddem, P. Renevey, E.M. Calvo, M. Lemay, K. Cox, D. Ashby, J. Willardson M. Bertschi, "Real-time Monitoring of Swimming Performance," Proceedings of the 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'16), Orlando, Florida, USA, pp. 4743-4746.
• Bertschi, P. Celka, R. Delgado-Gonzalo, M. Lemay, E.M. Calvo, O. Grossenbacher, P. Renevey, "Accurate Walking and Running Speed Estimation Using Wrist Inertial Data," Proceedings of the 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'15), Milano, Italy, August 25-29, 2015, pp. 8083-8086.
  DOI: 10.1109/EMBC.2015.7320269
• Delgado-Gonzalo, P. Celka, P. Renevey, S. Dasen, J. Solà, M. Bertschi, M. Lemay, "Physical Activity Profiling: Activity-Specific Step Counting and Energy Expenditure Models Using 3D Wrist Acceleration," Proceedings of the 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'15), Milano, Italy, August 25-29, 2015, pp. 8091-8094. 
  DOI: 10.1109/EMBC.2015.7320271
• Delgado-Gonzalo, E.M. Calvo, J. Solà, C. Lanting, M. Bertschi, M. Lemay, "Human Energy Expenditure Models: Beyond State-of-the-art Commercialized Embedded Algorithms," Proceedings of the 5th International Conference, DHM 2014, Held as Part of HCI International 2014, Heraklion, Crete, Greece, June 22-27, 2014, pp. 3-14.
  DOI: 10.1007/978-3-319-07725-3_1