Miniaturized Wireless IMU Enables Low-Cost Baseball Pitching Training Aid 1 Ryan S. McGinnis, 1 Noel C. Perkins, 2 Kevin King 1 University of Michigan, Ann Arbor, MI, USA 2 94fifty Sports Technologies, Dublin, OH, USA email: ryanmcg@umich.edu web: http://www-personal.umich.edu/~ryanmcg INTRODUCTION Baseball pitching is one of the most unforgiving positions in sports; one mistake, like a hung curveball or a fastball that tails out over the plate, often results in a run for the opposing team. As a result, there has been considerable research on the free flight behavior of a baseball, and specifically in identifying how a pitcher’s input to the ball (i.e. the ball’s velocity and angular velocity at release) causes it to break [1-3]. Experiments reveal that the total break of the ball during free flight is proportional to the aerodynamic lift coefficient of the ball [1], is dependent on the seam orientation [1], and is a function of the magnitude of the ball’s angular velocity [2]. These studies obtained their experimental data by tracking the position of the baseball in free flight using high speed video analysis systems. This method for data collection is expensive, time consuming, and requires an operator skilled in both the collection and analysis of the data. For these reasons, using high speed video analysis systems in baseball pitcher training is not an option for all but professional baseball programs. However, it is easy to imagine that an inexpensive and non-intrusive method for measuring the velocity, angular velocity, and orientation of the baseball would provide the foundation for an effective training aid for baseball pitching. To this end, we propose a highly miniaturized wireless inertial measurement unit embedded within a baseball (Figs. 1A-B) as a low cost, highly portable and minimally intrusive approach for measuring the baseball’s release conditions. Figure 1: The miniaturized wireless IMU (Fig.1A) is embedded within a baseball (Fig.1B). The black “tail” protruding from the baseball is a switch/recharging jack which, when removed prior to pitching, allows power to flow to the board. METHODS The inertial measurement unit (IMU, Fig. 1A), which provides three-axis sensing of linear acceleration and angular velocity, measures a mere 19 X 24 mm and weighs only 4.5 grams including a small lithium ion battery. The unit transmits data wirelessly using a proprietary RF protocol to a USB-enabled receiver which facilitates data collection on a host (laptop) computer via custom software. Subjects were instructed to pick the ball off of a tee, come to their set position on the mound, and then throw the ball to the catcher in an otherwise unencumbered manner. This sequence of events is readily identifiable in the example data shown in Fig. 2 where the ball motion is also divided into five distinct phases. Figure 2: Magnitude of the acceleration measured by the IMU embedded in the baseball for a typical trial. The 5 phases of the motion, as defined by our experimental protocol, are also indicated. Phase 1 corresponds to the ball being at rest in the tee (Accel. Magnitude of 1g) and ends once the ball is picked off the tee by the subject. Phase 2 begins when the subject picks the ball off the tee and ends when the set position is first reached. Phase 3 is the backward portion of the throwing motion, starting at A B