The measurement of the spatially distributed flow field in large industrial vessels, e.g. biogas fermenters, is not possible with commercially available measurement equipment. Therefore the concept of flow following sensor particles has been developed. They act as data loggers within the vessel. The sensor particles are equipped with a pressure sensor and an inertial measurement unit to measure their motion. The measurements are taken in the coordinate frame of the sensor particle and need to be transformed into the vessel's coordinate frame to analyze the flow field. This is done by an error-state Kalman filter which estimates the orientation of the sensor particle with respect to the vessel based on the direction of the gravitational acceleration and the local magnetic field. Since no other aiding sensors are available or usable within an industrial vessel, particular emphasis was given to the inertial sensors' stochastic modelling and the accelerometer's calibration. Based on the orientation estimation, the measured acceleration is transformed into the vessel frame. We performed two experiments in a 1.4 m3 lab reactor to qualify this approach. The results show, that the transformed acceleration is in good qualitative agreement with the known flow field inside the lab reactor.