Application of accelerometers in the sport of rowing

Rowing biomechanics is traditionally difficult to measure because of the nature of the sport. Rowing takes place in an aquatic environment, covers a relatively large distance (standard world championship race distance of 2,000 metres) and the mechanics of the rowing is complex (because of boat rigging, hydrodynamic interaction and the sliding seat). Traditionally, instrumented boats have been the standard for monitoring and analysing rowing technique, but they are expensive and complex to setup. Inertial sensors showed great potential as an alternative and were selected as focus of investigation. This thesis examines the use of Micro-Electro-Mechanical Systems (MEMS) accelerometers that are small, unobtrusive and relatively easy to set up, yet with the appropriate methodology can yield analogous information for rowing technique analysis. In undertaking the investigation of using a triaxial accelerometer as a rowing technique assessment tool, a thorough understanding of rowing biomechanics is required to help solve the inverse problem. One must understand how the shell acceleration trace is generated and how it relates to all the rowing mechanics variables in order to interpret it. Thus, the two aims of this thesis were a comprehensive rowing biomechanics model and solving the inverse problem to determine rower biomechanics using Micro-Electro-Mechanical Systems (MEMS) accelerometers. These aims were achieved and they were contributions to knowledge in the field of rowing biomechanics. The first contribution of this thesis was that it revealed the relationship between the combination of propulsion, resistance and rower motion against the resultant shell acceleration. This was achieved with the development of a rowing model to represent a single scull. The forces acting on the single scull and the resultant motion of the rowing shell was represented with a differential equation. A detailed multi-segment rower model was created to represent the rower motion. Also, a hydrodynamic model was developed to calculate the force at the oar blade, which is the propulsive force on the rowing system. On-water rowing data was collected and used as inputs to the rowing model to ‘simulate’ the rowing shell motion. The rowing model revealed how the rowing shell acceleration trace was generated from all the variables and parameters of the rowing system. The second contribution to knowledge of this research was the development of a methodology to use accelerometers with shell velocity and seat position measurements to monitor all the forces acting on a single scull and the resultant shell acceleration. The proposed methodology is based on a differential equation describing the motion of a single scull, which basically states that the force acting on the single scull is the sum of the force due to rower motion and the propulsive and resistive forces on the rowing system. The resultant force on the single scull was measured using a triaxial accelerometer, that is, product of the mass of the rowing system and the shell acceleration. The resistive force on the single scull was estimated from the shell velocity measurement and a coefficient representing the drag characteristics on the rowing system, that is, product of the drag coefficient and square of the shell velocity. The force due to the motion of the rower’s centre of mass relative to the rowing shell was estimated using seat position measurement and a compensation difference curve to account for the motion of the upper body, including the arms. The resultant force and resistive force on the single scull and the force due to the motion of the rower’s centre of mass were then used to calculate the propulsive force on the rowing system. The proposed methodology of calculating the propulsive force provides great insight to a rower's technique, as all the forces acting on the rowing system and the resultant shell motion are collectively monitored