How is it supposed to get a publicly routable address when it's not necessarily connected to the internet?
If this is truly disconnected from the Internet then yeah HTTPS is unsuitable - it fundamentally relies on the idea that there's a central, universal definition for who owns "foo.com" so that users can rely on talking to the correct "foo.com".