True story. I visited a pal's office and added his wifi to my phone and laptop to access internet from his place. A couple of weeks later, I had to visit his office again. Being a 'Google fan', I depended upon my Google Contacts to look up my pal's number and office address. That day, my SIM card failed and I don't know the exact building, floor and unit.
I took my chance and tried tower after tower, floor after floor. Suddenly my phone alerted me to an unread email! I'd got connected to my pal's wifi and he should be somewhere near. Using the internet, I contacted my pal and the rest's history.
What's the point of the story? It's existing technology that serves its primary original purpose that could be used for something else. A wifi AP is supposed to provide network access to wifi clients. However it can also be used for object locating.
How does it work?
Wifi AP or router has a limited range (distance) of effectiveness. Beyond that, either there's totally no association with it or the quality of network access is degraded. Very simply (there's other factors that I would not go into now), wifi is a candidate for near/far object locating.
Are there anything else?
Bluetooth is another hot favorite. There are a couple of recent tech pitches about using bluetooth stickers or devices to attach to assets for finding them. Probably there are already solutions using this technology previously but it's only recently tech news are featuring this.
Back to the past, there are already RFID solutions. IMHO, these are not readily available to the everyday us. We can't just go to the store and say, let's buy a couple of tags and then attach to our stuffs. Besides, mobile consumer devices do not have RFID capabilities. Pity.
NFC? Unless NFC can go further than just a bump...then it could be a potential candidate.
After all these, what is a missing gap to locate objects? Direction.
Besides near/far (or hot/cold), I haven't come across solutions that can tell me my object of interest in a particular direction. This is an important criteria, because if not, then we would be walking in circles figuring out where is the object though it's pretty nearby.
Is there a solution?
I believe there is. I'm not computer science trained, but besides common sense and chats with peers, there is a solution to this problem. I'm not sure how accurate or whether it would work. But got to try test this.
One, calibrate the locator (mobile smartphone). Move it in certain directions (like the figure of 8 for magnet calibration), collect data. What data? Read on.
Two, collect data from as many 'providers' as possible. They are the environmental noises e.g. nearby wifi APs. magnetic field data, GPS latitude/longtitude, altimeter, temperature, nearby bluetooth etc.
Three, data from known devices e.g. a mobile wifi/bluetooth device attached to the object via pairing/association and GPS check-in etc.
Four, create a virtual 'conceptual' map (2D/3D) to map the data collected.
The use case.
User attach a device (wifi/bluetooth) to the object. Calibrate smartphone (or locator). Move away from object and there should be corresponding information giving distance (near/far) and direction to the device depending on the movement.
Cool? I'm working it (android application). My first version would be a simple near/far proof-of-concept (PoC) and then add the directional function.
If you are interested and like to do something with me, feel free contact me.