Hello
I’m work on a project where it is necessary to detect if unit has moved and/or rotated. The unit (battery powered and wireless) is eg. placed on the floor and needs to be placed with same heading if it is picked up and put back up on floor within heading being +/-5-10 deg and in an upright position. The unit also have a PIR to detect movement in front of the unit. When repositioning and turning it to find the right heading, the unit will make a sound when the heading is found.
In the unit is a LIS3DH accelerometer and a LIS2MDL magnetometer.
From the accelerometer any movement will detected if the unit has been eg. picked up or pushed etc.
When put back the magnetometer is used to find the heading.
In test, it returns the correct heading each time, but can it be trusted over time?
True north etc is not needed since the heading is set once for this unique placement of the unit – it is more like a relative heading.
The accelerometer is always on, but magnetometer is off by default only if movement are detected by accelerometer the magnetometer is turned on. Also, magnetometer is shortly turned on each 10 minutes to verify the heading.
It is impossible to have user do a calibration like the 8-figure calibration so must be completely without. Can calibration be discarded since it is only matter finding the same heading again in the same environment?
If the position selected, will stationary metal which is close to the unit and the magnetometer is interfered, will the heading still be same when you put unit back?
What can possibly go wrong?
Thanks :)
