SparkFun Forums

[Modern Primate]

Hi all,

I'm hoping someone can point me in the right direction for a project I'm starting.

Concisely, what I need is to get a reading on a smart phone (or similar non-phone device like an iTouch or a tablet) about its location with a greater degree of accuracy than I've been able to get with the phones' internal GPS (I've worked with Android and iPhone). Preferably within a meter, but the cost/accuracy ratio would have to be taken into account.

In more detail, I work in educational technology, and in the past I've attempted to develop some location-based augmented reality games using Layar and the phones' internal GPS, but the technical problems were overbearing. I piloted one with college students (a very simple murder mystery that was little more than a few location triggered web pages), and many of the phones were unable to access the AR layer. Of the ones that were, many had difficulty triggering the points, even when they were right on top of them, due to the GPS inaccuracy. In the end, they ended up in four groups of five people sharing one phone when they were supposed to each have their own. The response was still very positive ("coolest thing ever," etc), but I wasn't comfortable trying it with actual kids, let alone doing something more complicated where the players would have an inventory and make choices that affect the game, etc, without figuring out a way around the accuracy issue. Even a better GPS would still be problematic indoors, which might be necessary in some circumstances.

Recently I had an ad on my sidebar for "WiFi based micro location beacons for iOs apps" but the page refreshed just as I was about to click. You'd think that would be a reasonably easy thing to Google, given that there was an ad and everything, but I haven't been able to find anything like this specific product, nor have I really seen a way to interface with it with the device in such a way that I get some coordinates or something I can run code against. I know retail stores are reportedly using WiFi tracking to see where customers go within the store, so I'm sure there must be a way to accomplish this.

Does anyone have any ideas about how I might be able to accomplish this reasonably cheaply? I understand that it would be helpful if I gave specific numbers, but really the best I can say at the moment is that the cheaper it is to do, the more likely I am to be able to raise the funds to do it and eventually get it into the hands of some kids. The devices themselves will be a pretty big cost. I'm open to any and all suggestions, including ones starting out "You're going about this all wrong..."

Thanks for your time!

[Mee_n_Mac]

http://en.wikipedia.org/wiki/Wi-Fi_positioning_system
http://dondodge.typepad.com/the_next_bi ... -maps.html

https://play.google.com/store/apps/deta ... paid&hl=en
http://www.navizon.com/
http://streetfightmag.com/2013/07/16/6- ... retailers/

[Modern Primate]

Thank you for the links, Mee_n_Mac. It looks like the most accurate of what you linked is Navizon, which claims 3 to 5 meters indoors. Looks like it would cost me about $1000 to run one game for Navizon (due to the monthly fee), plus whatever it costs for the devices. I wonder if it's concievable to pick up used iPhones on eBay and use this without a data plan. Should be, I think. I was hoping for something that didn't require a service fee, however, even if it was something I needed to build myself.

The trouble with 5 meters is that it gets you the room you're in, but locating something within the room might not work out so well. I guess I need to find out if there's a way to combine location based AR with visual AR so that the sprites actually stay on the ground without me having to use QR codes (or make the sprites impish fairies that float around "on purpose). Or maybe I should just stick with the QR codes and skip all this location based difficulty. Perhaps this technology isn't at the point I need it to be at yet.

[jremington]

You might take a look at what the SPAN lab headed by Neal Patwari at the University of Utah has been doing. They seem to be at a research forefront in passive indoor localization. Using a technique called radio tomographic imaging they showed how a network of cheap XBee radios could "see through walls" and quite accurately localize a human moving about within a building.
http://span.ece.utah.edu/radio-tomographic-imaging
http://span.ece.utah.edu/uploads/RTI_version_3.pdf
This recent review suggests that sub-meter resolution is possible:
http://www.cse.ust.hk/~yangzh/Yang-CSI-CSUR2014.pdf
The junior author of the SPAN lab paper has now formed a company called Xandem and offers a 15 node demo kit for $1500. http://www.xandem.com/

[Modern Primate]

Oh, wow. This looks very cool. I will need to think about how I can apply this to what I'm trying to do, or maybe change what I'm trying to do...

Thank you!

[bsagan]

As I think you are finding out, true positioning is not an easy task. GPS is good to a handful of meters; good enough for many applications like car GPS and for the most part, phone GPS. However, it's not good enough for the interior of buildings. I actually had a family friend who was investigating the viability of tracking the vector of footsteps to make a map of building interiors. I haven't followed up on the project for a while so I have no idea what ever happened to it. Find a brilliantly reliable and cheap way to do this and sell it to the military (they probably already have the tech but you never know haha).

[fll-freak]

GPS signals do not penetrate buildings very well at all. And if they do, they have so much multipath as to make it worthless. Local RF triangulation is possible but rather expensive and not in the skill level of most hobbyist. Optical systems work pretty well, but normally require significant processing and multiple cameras from several angles that can see the object to be tracked. Not great for multiple rooms. Sound waves can work sometimes but also suffer from multipath. A sonar shot into a room corner acts like a retro reflector and can give you multiple distance none of which are right. This gets down to IMUs. Current IMUs in the less than $1000 range can keep a position for perhaps as much as a minute IF the holder never exceeds the rates of the sensor. This can be greatly extended if somehow the holder can tell the IMU certain events. One such system zero out error each time a foot hits the ground. At that moment in time, the foot is stationary and the IMU and re-bias its gyros.

[carmen]

Check out indoo.rs - www.indoo.rs - which also appears in the Don Dodge articles and offers a solution that works on WiFi fingerprinting and/or BLE beacons. As far as I know, they offer some of the most competitive prices and even have a version you can use for free.

[Modern Primate]

Carmen, thank you. I checked that out and send them a message asking about the accuracy level.

Skye, yes, I'm getting that picture. I may have to either rethink what I'm trying to do or at least settle for a larger area outdoors.

[notexactly]

This system uses visual markers that are detected by the phone's camera. It looks to be quite good, but it's on Kickstarter and it's not doing very well funding-wise.

[Mee_n_Mac]

And another fledgling tech to look into ...
http://hackaday.com/2013/08/05/centimet ... s-for-500/
Where most GPS receivers only look at the data coming from the GPS satellites orbiting overhead, the Piksi uses another technique, real-time kinematics (RTK), to determine the receiver’s location with exacting precision. The basic idea behind RTK is to look at the carrier frequency of the GPS signals at 1575.42 MHz. This frequency has a wavelength of 19 cm, compared to the alternating 1s and 0s of the that are transmitted at around 1 MHz, or about 300 meters between each bit. While centimeter-level precision isn’t possible with only one receiver, two of these Piksi boards – one base station and one on a vehicle, connected via radio link – can make for a very exacting high-accuracy GPS receiver.

Previously, commercial RTK GPS systems have cost thousands of dollars – making a quadcopter or other homebrew project that relies on this level of precision nonsensical. [Colin] and [Fergus] have built hardware that can bring the price of this setup to under $1000. As a bonus, the Piksi board can also receive from other constellations such as Galileo and GLONASS. A very impressive piece of hardware, and we can’t wait to see the applications.

[fusedlightning]

To be clear with Mee_n_Mac's post, you would not be able to use RTK GPS with the GPS module inside any (that I know of) commercial smartphone. It would need to be added as an exterior module and interfaced.