I think a FHEM image for the neuron series is urgently necessary. My code is not yet optimal because it has no back channel.
Exactly the possibility in the control the position of the relays is interesting for many FHEM users.