A beyond boundary application. Could be used it to assess the quality of the rehabilitation process. As the author mentions, we have to analyze the application or use that we want to give the sensors and determine which data can become discarded.
Its me! http://www.bestbabyshower.co.uk/baby-shower-decorations/
While I take comfort in the fact that technology is moving forward in this area, it could become confusing for consumers in the long run as different platforms might have different ways of motion sensing, as well as different reconfigured actions and gestures.
Abby - http://www.comfortablefoot.com
This all comes to the use case. Optical systems with no instrumentation on the body (as Kinect)are efficient to a certain extent. When it comes to measure such information as as precise orientation of the hand or any limb, they do not fit very well. They also require the user to be in front of the system in a certain Motion Capture Volume. Probably a mixed approach sensor fusion is the top solution for these gaming peripherals.
When it comes to use cases where the user needs to get away from this Motion Capture Volume, you need him to "wear" some sensors.
Accelerometers are now very power efficient. Magnetometer and Gyros are more challenging but big progress are made every year as the MEMS industry (Gyros and Accels we use are MEMS, Magnos are not "Mechanical" devices but are manufactured also through MEMS process). So you can power a small Gyro, Accel, Magneto full system on a 100 mAh battery, with radio capabilities for a few hours before you need to recharge. This leads to a small node of a size of a wristwatch.
And its true that the number of sensors or modalities you need for an application depends 100% on the information you look for. You don't need a 3 sensor pack for all applications.
On the physical therapy domain, the analysis you propose is confirmed as these people are kind of experts in the fielf of human motion ! We delivered our kits to complanies in this domain, and now they are products derived from these kits, intended to the physical rehabilitation physicist. They use it to assess the quality of the rehabilitation process. And true that these same companies also target the assessment of sports athletes. The information they look for is somehow different than the one for the physicist, but can be obtained from Motion sensors and suited software. The number of nodes used is also growing according to how much you want a full assessment of the motion.
Yes, 100% Agree. The MEMS sensor toolbox is now very powerful but also complex to handle when it comes to deliver something which is derived from multiples modalities (eg Acceleros, Gyros, Magnetos, Pressure) and/or multiples nodes.
This is the part where we believe we can add our value. This is all about signal processing and sensor fusion.
I am wondering how much of the work could be done by external sensors aka Kinect? This would not have the limitations that a worn suit/sensor setup has and may provide more freedom of movement. Just a thought..
To reduce the intricacies of interfacing the MEMS sensors in a mobile application, it may be possible to offer a solution in the form of a sensor, its interface chip and the software driver. This will reduce the task of learning the interface techniques , right the rudimentary code to talk to the sensor etc. Senors with I2C like serial bus interface are most ideal to work with.
I wonder how power hungry are these kind of sensors?
And are all the accelerometers and gyroscopes MEMS?
Reducing sensors by developing a smart approach is a good formula, but I bet it will depend a lot on the application. As the author mentions, we have to analize the application or use that we want to give the sensors and determine which data can become discarded.
Good and interesting article nevertheless as it works as an introduction for the motion sensors. I hope we have new articles in the same topic.