Can you help me locate the guy who just stole my bike? We see this kind of thread regularly our Reader Forum: “Stolen S Works” (Dec 21, 2015); “Bike stolen en route to Cozumel” (Nov 27, 2015); “$1000 No Questions asked reward!” (Nov 12, 2015). “Bike stolen from apartment in Boston” (Nov 12, 2015). What we all want is an anti-theft device the size of a quarter and that requires almost no charging. I've got two technology solutions for you.


The first is TrackR. This Santa Barbara based company doesn't really make the product I'm looking for. It makes a nice solution for those who lose car keys, wallets, purses. Its device is so small it's unobtrusive. It can go on a keychain or attach to your dog's collar (dangling alongside his county license). Your lost item is detectable when it's within the BLE distance of a device carrying the TrackR app. Usually you misplace your car keys within the vicinity of your smart phone, so, no worries, your phone will locate your keys. (Just don't lose your phone.)

But it's better than that. The TrackR app finds your lost item when it's detectible by any device that's got the TrackR app installed. So, if you get home, find you don't have your wallet, and then ping your dog-tag-sized device, if that device in your wallet is within the BLE range of any device with the TrackR app installed (i.e., another smart phone), your wallet is discoverable. How is the location of your device determined? By CrowdGPS, that is, your search queries all the devices in the TrackR network, and when a device locates your lost item that device's GPS tells you pretty closely where your lost item is.

What you see when you search for your item is the last place the TrackR app saw your device, that is, maybe in the restaurant where you left it last night if one of the servers or customers had the TrackR app installed. TrackR has the technology to offer more than this. It can give you a history. A trail. This would be important for tracking a stolen item, because if your bike is spotted by the TrackR app 6 times over the last several days, and 3 are the same location, pretty good chance your item is there.

What if the dirtbag who stole your bike does not have the TrackR device on his cell phone, and your bike is in his house until he sells it on eBay? This is a problem.

One solution is to make sure the TrackR app is ubiquitous. Could the TrackR app be payload on a program almost everybody has? An internet browser. A mobile device's operating system. It will be interesting to see what TrackR does to speed up the deployment of its app. Some of this depends on whether TrackR sees itself as fulfilling the role of a LoJack, or if it's comfortable acting simply as a way for you to find your misplaced keys.

You might reply that this technology on a ubiquitious app exists. iBeacon and Eddystone are similar to TrackR, and Apple and Google do have ubiquitous placement on handheld devices. But these technologies don't work like TrackR, at least not yet. They don't crowdsource a search. They're simply ways for you to find the automotive section in Target, or your seat in a stadium. This isn't to say they couldn't compete with TrackR, they just don't.

The TrackR sells for $30, or 5 TrackR devices for $90. It's a one-time buy, no subscription, and the batteries are replaceable by the consumer. Tile is a competitive alternative to TrackR that works similarly.


There is another technology, it's not quite ready for prime time but it's getting there. Many of you have heard of the Internet of Things (IoT). A subset of the IoT is LPWAN (low powered wide area network). Imagine similarly small – dog-tag-sized – devices and these do more than simply provide location. Let's say you're a rancher, you have thousands of head of cattle over tens of thousands of acres of land. Would you like to know where your cattle are? Their body temperatures? Are you a farmer, you have hundreds of acres of fruit trees, you'd like to know what the temperature is throughout your orchard during a frost? As long as the packets of information are small and send out data packets occasionally, these devices can send out location, temperature, a water meter's monthly read, a pipeline's pressure, whatever metric you can imagine.

An LPWAN base station can be the size of a suitcase, perched someone up, as on a cell tower, and can exchange 1- or 2-way communication with extremely low-powered devices at ranges up to and beyond 10 miles. This is how the LPWAN works; the potential market is enormous; and there is a race on right now to build out networks, just as cellular companies have been engaged in their race.

What's even more exciting, the devices on your cows or trees – or your expensive bicycle – can get even lighter, cheaper and consume even less power because when an entire LPWAN network is built out your bike's tracking device won't even need a GPS. It can be located by triangulation of signals from several base stations, like the old pre-GPS LORAN navigational network.

There are two networks that get the most attention right now. One is called LoRa (not LORAN), the other SIGFOX. LoRa is a confederacy started by a for-profit company called Semtech. SIGFOX is a for-profit company. In each case the end-point (the dog-tag-sized thing attached to your cow) may well contain a GPS but instead of continuous monitoring of its location it might send out a packet (maybe just time + Lat/Lon) once every 5 or 10 minutes. It's extremely low power, and that kind of power use means the batteries will not need to be changed or charged for a year or more.

Devices on this kind of network communicate at a rate of 100 bits per second, which is slower by a factor of 1,000 than the networks that serve smartphones. The communication packets are small (12 bytes) and the data rate is slow (like, 300 baud). Because it communicates at such a slow rate it is easier for receivers to hear and understand the message, which is why you need way fewer – some companies advertise 1000 times fewer – stations to build out a LPWA network.

There are devices right now, which you can buy, that you can attach discretely to your bike, that will work. Just, as with TrackR, the question is whether the network is build out sufficiently so that your stolen bike will be located. It's got to be within range of of a LoRa or SIGFOX receiver. There is one company making devices with circuitry that can be located by both networks, but this is rare.

The beauty of these networks is that you need fewer base stations to complete a network than is the case with cellular. And, communication is over a sub-GHz unregulated frequency band. SIGFOX has France fairly well connected. The image below is a screenshot of SIGFOX'S deployment map. The network is fast getting deployed in many European countries. It promises to have 10 major U.S. metropolitan areas wired by April of 2016 (San Francisco, New York, Boston, Los Angeles, Chicago, Austin, Houston, Atlanta, Dallas, San Jose).

I get the value proposition for SIGFOX. It's the same business model as a cellular carrier. It makes its money selling subscriptions to device makers. LoRa is – if I understand it – more like a standard. Device makers are agreeing to a standard, so if I'm Cisco and I build base stations to the LoRa standard, some company building that dog tag transceiver knows any LoRa Alliance member is a candidate for base stations, and eventually a network, that will communicate with my transceiver.

To me, it's likely that LPWANs will be up and running within the next few years. If you live in a big city, you'll have networks already built out. Your bike theft device might even be already made, just not marketed as such. It might be a wayward pet finding device.

My one concern is whether it's just too early. The support structure for all of LPWAN is very thin. It's hard to find anyone to talk to. I remember teaching myself html code in 1993 and looking for anyone to explain to me the intricacies of the worldwide web. This feels like that.

In fact, it was reported almost 2 years ago that pet tracker Whistle is moving to LPWAN, but it still today uses a cellular network rather than a LPWA network, and it recently purchased Snaptracks, maker of competitor Tagg, which also uses a cellular network. So, we're close. But Whistle still can't deploy an LPWAN product because you might lose your pet outside of the tight confines of the urban areas in which these networks are currently deployed. If you point SIGFOX'S deployment map at the U.S., it looks sparse and bleak.

Bike Shop as Thief Detector

There are three places stolen bikes eventually show up: eBay, Craigslist, and the bike shop. Bike shops are important hubs. In each of the technologies above the bike shop can be a beacon of connectivity.

In the case of CrowdGPS solutions like TrackR, the bike shop personnel download the TrackR app onto their smartphones. We know the TrackR app installed on a smart phone identifies lost or stolen items when that item is within the BLE range of the phone. Imagine that bike shop employee getting an alert when an item reported to the TrackR network as lost or stolen was detected by his phone. A tone goes off. Let's all turn and look at the dirtbag who's rolling a stolen bike in the front door of the shop right now. I don't know that the TrackR app can be configured this way yet, but I cannot imagine it's more than a trivial task.

Let's say we rely on a LPWAN instead. The device we secrete in our bikes is picked up by LPWA network base stations. These base stations are relatively inexpensive. Bike shops might decide to purchase base stations, especially as these fall to affordable prices.

In either case, using either technology, bike shops would become a part of the anti-theft network, and part of the anti-theft solution. They would also be part of the commercial ecosystem by selling the devices that go into our bikes.