So, I've been using my http://blog.jgc.org/2012/03/ambient-bus-arrival-monitor-from... for well over 4 years and it's proved to be really useful. Today, I would likely follow Stavros' lead and use the ESP8266 since it's a nice module and OLED displays are cool.
What I learnt from the project was that the most important part was the model bus. This made it acceptable for my SO to have the thing in the house and she's the one that uses it a lot more than me.
I've built other "IoT" things around the house (lots of Amazon Dash buttons that don't do what they were intended for and other stuff) and the packaging matters a lot. It makes the things seem less like technology and more like part of the environment.
Since I wrote that blog post TfL have created a proper API and it has loads of goodies in it: https://api.tfl.gov.uk/
Perhaps unsurprisingly, I had the same reaction from my SO. She loves this bus, to my dismay, because my reaction is "it took me countless hours to build a sensor and automation network for the house so it can all be centrally controlled, and you like a two-hour bus build?"
It seems that people like things by how useful they are for them, not by how cool they were to build, which is frustrating but not surprising.
The packaging is why I got a 3D printer and am learning CAD. I want to create little stylish boxes for my sensors, so they look more like devices and less like bombs.
That's true, although the motion-activated lights work well enough. I think it's just the utility for her, because she can just glance at the bus and know when she needs to leave, when it's not that big a hassle to flick the switch when you walk into a room.
I don't know what JGC did, but you can detect the button's DHCP request and consider it pressed. I didn't want to do something that hacky, so I made my own Dash:
Wow, an excellent post! I'm really motivated now to finish my own project, which is a little bit more bulky (a lot more bulky: An old 7" netbook which has been gutted and I made a wooden frame and box to hold it and put it on the wall). I'm still thinking of what to do with it, but I'd like it to be sort of a hub of info, with time, weather, news, HN, bus, etc., which me and other people in the house can use. I'd also like to implement a webcam feature which would turn off the display unless there is someone near the device moving.
The takeaway for me is that it costs roughly $15-20 of hardware parts to have a stand-alone status indicator. That's pretty cheap, but not "no brainer" cheap.
I start asking myself what are the status indicators we really need? That people would pay $50 for?
Is there any alternative to wired USB power for these kinds of projects? Are there any teeny wifi boards that can be powered for long periods by a AAA, 9V, or button cell?
"Long periods" "wifi" and AAA, pick two. Wifi is too power-hungry for batteries, unfortunately. You'd probably get a day or two of continuous operation on a biggish battery, I'm guessing.
You could probably use Zigbee for something like this, although I don’t know whether there are any v.small zigbee capable arduino interfaces. Running the display is going to take some power too, unless you can find a small e-ink option.
Can you make small pieces of paper stick to a battery? No. Can you make them stick to a balloon if you rub it on your sweater? Yes. Greece - Italy 1-0.
What I learnt from the project was that the most important part was the model bus. This made it acceptable for my SO to have the thing in the house and she's the one that uses it a lot more than me.
I've built other "IoT" things around the house (lots of Amazon Dash buttons that don't do what they were intended for and other stuff) and the packaging matters a lot. It makes the things seem less like technology and more like part of the environment.
Since I wrote that blog post TfL have created a proper API and it has loads of goodies in it: https://api.tfl.gov.uk/