Performance monitoring for professional and recreation sports using wireless sensor networks

Nadia Khaled,  PhD, EPFL-IC-LSI

To cover the wide range of competences needed in the project, we propose a multi-disciplinary team of partners coming from four different countries: The Netherlands, Spain, Italy and Switzerland. The following two labs from Centre SI are the partners representing EPFL.

Integrated Systems Laboratory (LSI – Dr. Nadia Khaled) will act as the system architect and integrator. Its research will focus on designing energy-efficient air interfaces and MAC protocols for the target wireless sensor networks, power management strategies for wireless sensor nodes and low-power processing architecture for the sensor nodes.

Microsystems Laboratory (LMIS2 – Prof. Martin Gijs) would provide both vertical and shear force sensors.

The achievement of high-level athletic performance is increasingly associated to careful monitoring of all key metabolic functions, both during the effort and during the recuperation phases. Such key metabolic functions include, but are not limited to, heart and respiratory rates, blood oxygen and carbon dioxide levels, glucose and lactate concentrations in the blood. To acquire and monitor these metabolic functions in a non-intrusive fashion and in realistic outdoor training and competition set-ups, it is highly desirable to design an innovative, low-weight, acceleration-resistant and autonomous wireless sensor network system. Such a wireless sensor network should be capable of acquiring the desired biometric information, process it and communicate it to the coaches, trainers and medical staff, who would use this data to accurately assess the athlete’s in-situ physical condition, energy expenditure, and accordingly devise appropriate training and diet programs, as well as take optimal tactical decisions. In a nutshell, the complete system would contribute to better training supervision and tailoring, reducing the athlete’s over-training and ultimately improving athletic performance. It is anticipated that the technology developed within this project for high-level athletes will find wide acceptance among recreation sports fans, and sedentary people following a weight loss program.

Research Goals:
A detailed state-of-the-art overview and our interaction with exercise physiologists showed that most current wearable sensor platforms are bulky, intrusive and mostly wired. Furthermore, the current few wireless sensor platforms are largely power inefficient, which seriously compromises their autonomy and usefulness. Consequently, the challenges of this project are manifold:

  • Devising miniature and low-power sensors, equipped with on-board ultra-low-power processing to preprocess the acquired data such that power-hungry wireless transmission and reception are optimally reduced.
  • Jointly devising and optimizing energy-efficient air interfaces (PHY) and medium access control (MAC) and networking protocols, which target maximum wireless sensor network life span.
  • Developing and enhancing advanced electronic integration technology to enable integration of sensor nodes on flexible substrate and stretchable structures. This will achieve comfort for the sportive by enabling integration into clothing or clothing accessories. This will in turn favor acceptance of the system by the end-users, known as a major challenge of Body Area Networks.
  • Exploration and optimization of an application/service architecture for the ultra-low-power processors on-board the developed sensors.
  • Realization of tactile force sensors for measuring both normal and shear forces of the hand: the sensors will consist of a layer of planar Cu coils combined with an elastic polymer layer, both of which are sandwiched by two magnetic foils of high-permeability material. The measured inductance is a very sensitive function of the magnetic gap formed by the thickness of the elastic foil, so very small deformations of the elastic material, and, after calibration, forces can be measured.