When I was living in a residence hall a few years ago, I envisioned a device that could act as a kind of smart front door intercom. It would let my friends who came by my room know if I was in/out, available/busy/sleeping, etc., let us hear each other through the thick door, and let them leave messages or call my phone from my door if I was out. I then thought it would be good to have a touchscreen on the inside of the door, to display my calendar, the bus schedule/transit trip planner, etc., as well as an automatic door lock so I could lock/unlock my room with my phone, with my student ID, or by face recognition.
I'm no longer living in a residence hall, but I am now a member of a hackerspace, and I've realized that a very similar device would be quite useful for the hackerspace's door. Currently we only have a Raspberry Pi 1 that does simple RFID access control and runs a Twitter feed that just says if anyone's in the space or not. I would like to build a new device to take over these functions as well as provide several other functions, such as a security camera, automatic lighting of the door area when someone enters, warning people not to enter when someone is e.g. welding right behind the door (as well as notifying the welder to pause), taking messages from non-members who come by when nobody's in, giving information on nearby buses that are leaving soon, etc.
I intend to develop this system with an eye toward making it easily adaptable to other applications including residential use. I was originally planning to use a Raspberry Pi 1 (before the 2 came out), so the Raspberry Pi 2 is a natural fit. It should be easy to interface it with the other hardware I'll use (cameras, lights, RFID reader, motors, speakers/microphones, etc.). I expect/hope Windows 10 will make the development easy (though I have no previous Windows development experience). Microsoft Azure could be used to provide a monitored burglar alarm/remote premises monitoring service, track resource usage, or host our members database (though several members have an aversion to having their data stored in the cloud), or its machine learning feature could be used to implement a smart thermostat that predicts how many people will be in the space at different times of day and controls the furnace appropriately. Residential applications could also benefit from most of these uses of Azure, as well as others such as grocery list automation and sharing.