Okay, so I was thinking about “house logs for dummies,” and one thing became generally evident. For virtually every project that you do, the step one is identical: survey. Building a new home? Survey the land before laying the foundaton. Moving into a prebuilt home? Survey the prebuilt structure that is lacking the blueprints. Trying to get data on a home in operation? Survey the home so you can determine how you need to instrument the sensor network. Detecting inaccurate data caused by missing data in your sensor results? Grow a bigger sensor network, but of course sensor network growth must be guided by… yep, you guessed it, survey data.
Whether the project be to build something out or to measure something in operation, the universal first step is to survey. Unfortunately, I don’t think I can say there is a universal second step. If you want to run things in a rich data-driven manner, the second step would be to build out a sensor network. But, if you want to do things bare-bones, you could just as well forgoe surveying and data collection until you need to perform a major undertaking. Yes, this does mean that if anything goes even slightly wrong, you risk forgoing the ability to catch the problem early, rendering what might be a small fix into a major repair.