Motion Tracking Using Inertial Sensors -- COMPAMED Trade Fair

Motion Tracking Using Inertial Sensors

Photo: Matches in a box

If you like to measure motion, like for instance the motion of a walking human being, it is preferable to require little equipment and to be able to collect data that can be quickly analyzed with the available devices. Especially image-based gait analysis was not always able to meet these two requirements, which is why research headed for the so-called inertial sensor based gait analysis, as it is for example performed at the Fraunhofer Institute for Factory Operation and Automation. There, a system was developed that on the one hand is able to identify specific diseases of the musculoskeletal system, but is also meant to support treatment after stroke or cerebral tumor.

Graduate engineer Martin Woitag, a research associate on the Fraunhofer Project, explains: “The advantage of inertial sensors is that you don’t have to capture the movements with cameras. There are camera based motion tracking systems, where a marker is pasted on different objects and these are then captured with the camera. This is not the case with inertial sensors, you don’t need any references all around, but instead you attach the sensor directly to the moving object. This in turn sends data for position and orientation to a receiver. “The sensor itself, one per foot, barely is the size of a matchbox and therefore doesn’t bother the carrier.

Inertial sensors are small and easy to attach to the object

The foot motions are captured using different acceleration- and rotation rate sensors. The collected data can then be chronologically and spatially related to one another. Woitag explains this in more detail: “We primarily work with so-called acceleration sensors, which measure through validation in three different axles and with rotation rate sensors, which measure the angular rate around an axis. From this collected data, you can then deduce orientations and positions in space, for example where a foot is located in space. “


Photo: Feet in motion


This collected data on the one hand can be used to check the gait symmetry, but can also on the other hand be used to check treatment during rehabilitation for example. But completely different applications are also possible with inertial sensors. At the moment, Martin Woitag is looking into the possibility of using inertial sensors for automatic time-recording of assembly processes – because the data the sensor measures will determine the application. The objective is for instance to determine the best positions of work material. Unimportant positions can be excluded: “You conduct a training phase, during which the system is taught different positions. This way, the measuring system learns which points it needs to capture and which ones it shouldn’t capture. Uneffective grasping times are thus being eliminated. From the data for position determined during the first step, the time data is then deduced during the second step, since the sensors supply a position for every point in time, and as soon as a specific position was captured, you can time the interim. This way you can measure the total time, like for instance how long a grasping movement lasts. “The ideal motion and length of time is thus determinable thanks to the data – unimportant motions like for instance scratching one’s nose are automatically deducted.

Better prosthesis for leg amputees

Inertial sensors are also soon to be used more extensively in other areas of medical technology – for example in the further development of artificial legs. Just at the end of last year the research project “Measurement-based gait optimization“ was awarded the Medical Technology Innovation Prize 2010. The goal: inertial sensors in artificial legs are to deliver data on rotary movements, acceleration or slants of the prosthesis segments. Professor Marc Kraft, the manager of this project, states in a press release: “We want to design a mobile gait analysis, this being a measurement technique that is integrated into the prosthesis, so that the patient’s movement data cannot just be measured under laboratory conditions, but also in real, everyday situations and at the medical supply and equipment store where the patient is being taken care of.“ Prosthesis wearers therefore can now hope to receive better prostheses that meet their needs in the near future.

Simone Ernst

(Translated by Elena O'Meara)