I ‘d like to build a high accuracy pedometer/distance meter not bases on accelerometer, but rather two distance sensors placed is boots.

Both boots have receiver and transmitter and distance betweem boots is continously measured and added to overal steps.

Of course some reasonable value bases on propper filtering has to be calculated.

However, i can’t use optical sensors since it all should be placed inside boots. Also an ultrasonic probbably doesn’t work.
So prob. best option is magnetic sensing with halls or coils. What do you think?

 

Building an accurate pedometer which just records the number of steps sounds possible as each step would bring the sensing devices close to each other. A magnet clippe to one boot with a hall sensor clipped to the other would trigger when the feet pass each other and a realistic debounce delay should eliminate over counting

But magnetic field falls away according to the cube of the distance away so even a strong magnetic wound have too small a field to be a reliable measure the length of a step which could be 80 cm or more.

Ultrasonic sensing is good for sensing small distances more in line with step distance. Mounted on ankle straps would make it more flexible. Whilst ultrasonic distance sensors designed to work with an Arduino are cheap to buy I think you would need a custom arrangement where the device on each ankle is actively “talking” to the other rather than the usual measurement away from a passive object.

Good luck with this! An interesting idea

 

Welcome to AAC.

Why are you trying to avoid accelerometers?

 

Thank Jerry.
I was thinking to use a simillar approach like metal detectors (Rx,Tx coils), but rather than sensing reflection sense a passed field thru Rx coil where a matched LC tank is placed. The coils can be wounded inside boots, Tx on tip and Rx on back ( or vice versa).

To Yakov:
I think a higher accuracy should be achieved with direct distance measurement than running algorithm to get distance from accelerometer datas. Maybe not, who knows. But certainly it has to be think the every kind of step (fast, slow,running) is so individual/ unique.

 

To Yakov:
I think a higher accuracy should be achieved with direct distance measurement than running algorithm to get distance from accelerometer datas. Maybe not, who knows. But certainly it has to be think the every kind of step (fast, slow,running) is so individual/ unique.

Given the complexity of the processing you would have to do with some sort of measuring between the feet, and that people will spend time in the air as well, which means that you won’t be able to directly tell how far their running stride actually took them, using an IMU and sensor fusion along with clever computing seems much more likely to result in a high accuracy measurement.

Inertial methods could determine how far the person has gone based on both the acceleration and impacts that would also give you step count. There are so many IMU options today and some sort of GPS-disciplined inertial measured or other clever combination seems (to me) far more promising then trying to measure the distance between feet and fudge the airtiime you can’t measure directly into some number.