I am updating the building foot prints layer manually by tracing from the most recent aerial imagery tileset available.
There is probably a better way of doing this than digitising it manually. This is just the method I've been instructed to do it. ¯\_(ツ)_/¯
Hawkesbury is a really big area and if we go in to update the dataset based on aerial imagery manually, we can't place confidence on the dataset after 3-5 years due to changes to urban areas. The continental drift will also be a problem over a longer period of time. I assume surveyors capture it with LIDAR with a plane or drone, then extract the building footprints from a point cloud cluster.
In the five years i've been here there is only one report that requires the use of the dataset. A state government agency has an annual request for information, largely for reporting purposes, to determine the number of buildings added or subtracted from the area, and then how many buildings are deemed to be affected by flooding. But the definitions they use to define the reporting area is misleading, so I bet everyone fudges the numbers to get it done. No one ever comes back to verify the data we submit so it sounds like some arbitrary figure that is reported to the agency every year and no one checks it? :D
If the planners or enginers wanted building footprints, you could survey the area with LIDAR or a drone, and then capture the study area in its entirety. But I wouldn't maintain a massive dataset of building footprints without seriously consulting experts about how we could automate the process.
Amusing Insight From Aerial Imagery
Every now and then you see something interesting on aerial photography that wouldn't be evident from being on the ground. This could include unusual geological formations, patterns of urban development or a geoglyph of a penis, that can be seen from space.
Upon inspection of our jurisdiction, I came up with these two below: