My house had a not insulated attic / roof, because the underside of the roof is directly visible it is easy to inspect for potential leaks. I was able to catch one fairly early several years ago.
Now I plan to install a heat pump in my home and one of the first things HVAC guy advised was to insulate the roof with a thick layer of rockwool.
I was thinking to add some sensors on the underside of the roof before I install the rockwool / damp barrier in order to detect potential leak issues in future.
What do you generally think about such idea? It would be a really long term installation (20 years?) is this even feasable with our current sensors technology and what sensors would you suggest for such install?
Is your attic going to be used by people? Why insulate the roof if no one is going to live there. Insulate the attic floor to stop heat loss from the main part of the house. Is it not already insulated?
Around here, you want the attic roof to be not insulated so the snow will not melt quickly and slide off the roof and take the rain gutters with it, or form dams at the edge of the roof and store water that will leak into the attic.
What sort of sensors are you considering? You would need thousands of moisture sensors to watch for water under the roof.
It's a bit complicated, there are many factors that favor insulation of the attic roof. I want to store things there without them cooking in summer and freezing in winter, we also have holes in ceiling to let in light via roof windows into lower floors etc.
All in all it makes sense to insulate the roof instead of improving already existing insulation in attic floor.
Regarding detection, I just remembered about my datacenter and water detection cable under the double floor... It is very trigger happy, does not need to be submerged in water, simple water spill close by will cause trigger. I remember it triggered when cleaning lady walked in and washed the floor with a mop....
It costs about 5$ per meter for non-positional wire, so I'd need ~1.5k for the wire alone. It's expensive but can save a lot of trouble if it detects leaks early on.
I think you'd need a network of humidity sensors and monitor them for unusual high values appearing over time - once an area gets moisture in the humidity will tend to rise in the vacinity, but humidity will have a large daily variation too due to temperature changes, so automatic detection isn't straightforward I fear.