We're considering using a turnkey video surveillance / time-lapse archive service for on-demand surveillance of a small construction project (about 30 acres). Since there is little in-house experience in this area, it seems to make sense to pay a premium for an all-in-one solution, instead of accepting the risk of trying to run it in-house.
Even so, I need to perform the due diligence to ensure we won't pay an excessive fee, or get stuck with something unusable, when we might have been able to put the pieces together manually.

It looks like 4 or 5 cameras should provide comprehensive visibility.

So, the purpose of this post is to get feedback about using a service solution provider (e.g., Earthcam, Oxblue) versus an in-house configuration. I also have a couple general questions about camera capabilities.

Any feedback on these points will be greatly appreciated:

- It appears the trend for communication is to connect each camera directly to the comm carrier via 4G, instead of connecting them to an intermediate on-site controller first [?]. If you only have a few cameras, it would seem ok to pay for the data plan / hosting per-camera. But at some point (e.g., say you had 16 cameras), you'd want to connect them to an on-site controller (not necessarily a recorder), then connect the controller to the ISP -- right?

- Camera technology seems to be changing rapidly [?]. Will cameras purchased today be worth using / maintaining in one or two years?

- We have fulltime IT resource. How complex (or simple?) is it to host a VMS system? I see where AWS has a couple marketplace offerings...

- The consumer-grade camera market (Arlo, Ring, etc.) offers hosting with the camera. Is that true of the commercial market too? I guess what's being asked here is why pay the premium for a service provider's VMS, if you can use the one that comes with a purchased camera.

- Any camera have a screen wiper or blower for raindrops?

Thanks very much for your patience, and for any feedback you can provide.