In March, a man was shot and killed within view of a surveillance camera in a Miami public housing community. It was not untilpolice tried to pull video that they learned the camera was not working. The wiring had been ripped out.
All too often, CCTV is deployed to "tick a box" but there is far too little effort to ensure a system is operating as initially intended over the life of the system. Automated remote health diagnostics should be included as part of the solution for these types of systems.
I would guess someone knew. There was only ~20 cameras so it is not as if it would be that hard to tell. It appears from the city's actions afterwards that their was no maintenance contract in place, so even if they did know, they did not have the money to pay for the fix.
Unfortunately, I have heard too many "safe city" customers say that their number one goal is to use cameras as a crime avoidance measure. The fact that cameras are there should be "enough" I have always explained the value of higher resolution (partially so that you can mount them higher) as well as continuous recording instead of motion (to capture false positives). When confronted about budget, I tell users to reduce camera count in favor of quality. The public can get mad about wanting more cameras, but likely won't sue you if a camera doesn't exist. Perhaps Miami should use the Lancaster model and adopt cameras ;-)