So I am a physics student on a 2 month internship with a small company working as a solo research and development team. The company makes washer disinfectors which clean toiletry utensils for places like hospitals and care homes. They gave a a very broad design brief to incorporate more technology into their machines which are mainly controller by fixed circuit-boards and relays and stuff (I don't really know circuitry well). So I was told they were looking into using a raspberry pi (model 3 b) to control their whole washing cycle so I have been learning linux, node-red, javascript to help me use it ( I previously only knew python).
I have made the node red flow and it talks to a relay module that I have purchased and basically at different points in the cycle it will turn the relays on and off to control hardware like pumps, steam generators and door mechanisms.
So I would like to gather ideas about how i could set up a flow which tests each of the components and what hardware I would need to do it. Because obviously I cant get any data back from the components back up the relays, I assume I would need to make like ammeters and test them that way?
Anyway, I'm not so sure where to start so I thought no harm in asking here. Sorry for the long essay I thought the more detail you had the easier it would be to get answers.
Hi
Interesting project ! There is some detail in your question and it looks like you have made some progress, but there are very many different ideas and possible solutions. I think that the control functions (relay operation, timings and core operation) is ideal for Node Red, the other parts of you question about hardware etc may not be suitable for this forum (but there may be many people here who could help).
Ideas and Suggestions
1.) You could make it cheaper per unit by not using a Raspberry Pi 3 for each device but using a cheaper ESP8266 device (Wemos D1 Mini) and then connecting feedback from each component to determine failures. These would communicate by Wifi and MQTT to a central Raspberry Pi which would run Node Red and have the functional flow.
2.) The above example could be connected by MQTT to a cloud based instance of Node Red and you could monitor all of your installations over the internet from the office
This may sound complex but it is certainly achievable
Good luck
I agree with @mtoko about using inexpensive Wemos D1 Minis as your slaves.
If you flashed them with ESP-Easy you could arrange for a GPIO output pin (that drives your relay) to be sensed and fed back (using ESP-rules) to Node-RED via MQTT. This means it closes the loop and provides a certain amount of feedback.
Thank you all, I'll have a look at those methods. I had a discussion with my manager and I mentioned the cost effectiveness of having each of the components telling that its not working. Say if the pump stopped working it would only be a matter of time before the user calls up about a problem. Plus its most likely that care homes having reliable internet or internet period to see such functions.
We settled with a cycle counter that triggers a message to be displayed after so many cycles to call to get a service, That is inexpensive and will help the company greatly.
Anyway thanks for the help, Node-Red Forums have been really quick and friendly with replies, such a nice community.
