Replace magnetic connections with small, self-aligning, cyclable electromechanical connections #2
Description
The magnetic connections are a a clever idea due to their self aligning nature basically infinite use cycles. However, they introduce a major issue due to their proximity to the magnetometer. Alone they introduce a strong local field that must be modeled and subtracted in order to see the weak geomagnetic field. This is not ideal, but also not insurmountable on its own. The real issue is when the device is used in a multi-animal context. The relative motion means that the other's animals' magnetics are not static and thus cannot be distinguished from true motion relative to the geomagnetic field during sensor fusion. This results in spurious orientation readings.
This screen grab shows the fused Quaternion from of a stationary device as I move another optogrid around it:
