That’s a noble idea, but we need to keep the stack tight. First, make sure you can route packets with low latency—use a dedicated VLAN for the hub so the shelters don’t interfere with your own traffic. Put a firewall in front, and think of authentication like a handshake: no rogue devices, just captive‑portal‑authenticated clients. Add fail‑over with a spare router and a static route table so a single outage won’t knock the whole thing down. And don’t rely on a “smart” cloud gateway you didn’t set up; write the config yourself so you know every rule. If you can keep the topology simple and monitor logs in real time, the hub will stay online long enough to help.