Sentera’s strength lies in the integration of sensor elements with a signal conditioning and processing unit. Depending on the complexity and power constraints of the application, the required digital filters and algorithms are embedded in a microcontroller and/or digital signal processor. Sentera's library of signal processing and motion-specific algorithms have been implemented and optimized for a variety of micro-processors that range from ultra-low power microcontrollers to floating-point digital signal processors.

Implementation of motion sensing algorithms varies substantially from product to product. This variation is caused by the fact that algorithms used to interpret the sensor signal can only estimate the physical reality of the motion. Algorithms can significantly improve the interpretation of the captured signal by applying judgments as to the relative importance and validity of the motion signal in the context of the specific application.

An example of this is a vertical reference, used to measure dynamic roll and pitch angles. Accelerometers are capable of measuring roll and pitch angles but are only useful in limited conditions. Lateral motion adds "noise" to the accelerometer measurements and therefore accelerometers are only accurate in measuring long-term angles. Gyros, on the other hand, measure the rate of rotation and are unaffected by lateral motion. Gyros however suffer from long term drift that cause the angle measurement errors to grow with time. The solution is to complement the short term accuracy of the gyros with the long term stability of the accelerometers to achieve the optimal estimation of the dynamic roll and pitch angles.

The fusion of accelerometer and gyro data is performed with a Kalman Filter. A Kalman gain matrix is computed according to parameters such as sensor-specific noise parameters and the current sensed lateral accelerations. Computed with the estimated error characteristics, the gain matrix determines the optimal combination of accelerometer and gyro measurements. After each Kalman update, the covariance matrix is updated to reflect the new error state and is then used in the next iteration. In this way, information about the current state’s error characteristics is used to determine the optimal use of the reference measurements.