Inertial Motion Sensors

So, we now know that wearable sensors are the most reasonable choice for capturing our motion performances. And while there might be several wearable devices on the market out there, we now just want to talk about the most common ones – inertial sensors. But what are they all about?

Inertial sensors offer many features that are ideal for mobile motion capture and tracking applications: they are cheap, light and capture human motion in an easy, flexible and direct way. In general, they consist of a combination of miniature three-axial accelerometers, gyro rate and magnetic field sensors. Those sensors supply the user with information about gravity and external acceleration, angular velocity and the magnetic field. In a next step, they can then be processed to provide more intuitive and meaningful motion data. To start, the different built-in electromechanical sensor device types are described in more detail.


An accelerometer measures static and dynamic acceleration forces. While gravity causes a constant static acceleration, moving or vibration of the accelerometer causes a variate, external dynamic acceleration.  By measuring the amount of static, gravity-induced acceleration, one can determine the angle the device is tilted at with respect to the earth. By sensing the amount of dynamic acceleration, one can analyze the way the device is moving. To get a 3-dimensional information, inertial sensors usually combine three perpendicular accelerometers. They define the axes of the internal sensor coordinate system.

inertial sensor acceleration


A gyroscope measures angular velocity and rotational forces. It consists of a wheel or disk mounted onto a spinning axis in the center of a larger outer ring called gimbal. The axis of rotation of the inner rotor is free to assume any orientation by itself. As the outer mounting is tilted or rotated, the rotor remains stationary and maintains its direction in space to conserve angular momentum. The outer rotation can then be annotated and the respective angular velocity be registered. To determine the angular velocity along three dimensions, a gyroscope can be equipped with three gimbals.




A magnetometer measures the strength of the magnetic field at a point in space. Aligned along three perpendicular axes to a coordinate system, the effect of device rotation relative to the magnetic north can be measured and the orientation of a sensor device relative to the Earth’s magnetic north be determined.

magnetic field

Types of Inertial Sensor Devices

Technology for full body motion capturing of sport performances was introduced commercially at the beginning of the new century. Many research facilities all over the world use the commercial X-Sens system (XSens Technologies B.V. MVN/MTw, Enschede, Netherlands) for their research, and it serves as a quasi-standard since it offers high accuracy and usability. The included data processing and graphical user interface software are further reasons that make the devices popular among researchers worldwide. However, the system has a relatively high initial cost. The sensors are furthermore of relatively big size and not waterproof. Here, alternative hardware solutions can often offer a cheaper, smaller and waterproof combination of inertial sensor modules.

In recent years, more and more several independent sensor modules became available. For the majority of the applications described on this website, I used 9-axial measurement units from Logical Product (Logical Product. SS-WS1215/SS-WS1216, Fukuoka, Japan). They contain triads of gyroscopes, accelerometer and magnetometer for the respective x,y and z axes as visualized in the following figure.

inertial sensors principle

The biggest problem of independent sensor systems is that they usually do not evaluate full body kinematics. This means that they only offer raw sensor output data that cannot represent positional or angular information. To enable a meaningful full-body performance analysis, these properties need to be derived specifically, whereas the sensor data processing is generally independent of the device used.