I have built a system to look after my vivarium runing on a Mega 2560.
It monitors temperature and uses PWM to control a ceramic heater using the 'dimmer' method, can fire a humidifier, cooling systems, turn lights on and off etc etc.
I have written software that runs on my phone and connects in via Bluetooth to allow reading and writing of settings from/to EEPROM.
So far so good.
The next step I am considering is allowing access to it over the internet via a web browser or dedicated app and here is where my problem starts.
IF I was to make this a commercial venture I would want to make absolutely sure it cannot be hacked (Yeah I know - there's always a possibility...).
So I was wondering if it's possible to flick a virtual switch which makes the board read only and then (probably using something like Google Authenticator) if I ever need to do an update I can switch it back to read/write mode.
I have all sorts of checks in place to raise alarms and turn down power in the event of an emergency but of course this could be changed if I got hacked.
I can of course build a seperate system to monitor the main one that would not be internet enabled and could throw a kill switch but I'm hoping to avoid that.
Thanks for any replies!