Universal algorithm for exercise rate estimation in walking, cycling and rowing using triaxial accelerometry

A technique that can reliably monitor exercise intensity plays an important role for the effectiveness and safety of an exercise prescription. A universal algorithm for the recursive estimation of exercise rate during a variety of aerobic exercises using measurements from a body-mounted triaxial accelerometer (TA) is proposed. Information about the type of exercise is not required by the algorithm and the TA can be mounted at the same location regardless of the exercise type. The algorithm involves period detection and data fusion. Experimental results demonstrate that the algorithm is effective for common aerobic exercises.