No Map? No GPS? No Problem
Startup Navisens says it can find people indoors using motion sensors and math.
Now that it’s easy to find your way in the real world with just a smartphone in hand, the next logical navigation frontier is indoors, where GPS doesn’t work and maps are often nonexistent. Australian startup Navisens says it has a plan to track everyone from firefighters searching through burning buildings to consumers wandering through shopping malls, without requiring any special wireless signals.
Ashod Donikian, Navisens’s founder and CEO (and, for now, its sole employee), says the company’s technology will power a wearable gadget that can track first responders in emergencies; he hopes to start testing it shortly with firefighters in the United States. He also plans to offer his technology later this year to developers who want to add indoor positioning capabilities to smartphone apps or enhance existing navigation capabilities.
Navisens is one of many companies trying to make indoor location tracking possible (see “The Indoor Positioning System Era”). Common methods for indoor location tracking include using Wi-Fi hotspots or cell-tower locations, but these aren’t exceptionally accurate and usually depend on existing wireless infrastructure. Some more creative tracking and mapping methods are also in the works (see “Startup Uses a Smartphone Compass to Track People Indoors” and “Smartphone App Crowdsources Indoor Floor Plans”).
During our meeting at a San Francisco coffee shop, Donikian shows me a prototype of the gadget he hopes firefighters will soon strap on to their gear. It’s about the size and shape of a travel soap case, painted the color of a Granny Smith apple.
Inside, an inertial measurement unit—composed of three accelerometers and three gyroscopes mounted on three different axes—captures information about its acceleration and angular rotation, and a small computer uses algorithms to determine location, current direction, and the wearer’s position—standing or lying down. A recorded path is wirelessly transmitted to a USB dongle connected to a remote computer, where another person would be monitoring the wearer’s movements (often, this monitoring is currently done by radio). Donikian calls the position-calculating part of the technology motionDNA.
Navisens uses dead reckoning, meaning current location is calculated from previous location and movement—in the case of the firefighter, the fire truck would serve as a starting point. While this method becomes increasingly inaccurate over time, Donikian says, the algorithms Navisens uses cut down on errors.
“The key here is we know where the person is in the building well enough to send a search party to narrow down and find where they are, and we’re doing this without placing any infrastructure in the building,” he says.
Jim Duckworth, an associate professor of electrical and computer engineering at Worcester Polytechnic Institute whose research includes a first-responder tracking system, likes the idea of a navigation tool that requires no infrastructure or prior knowledge of the building you’re entering, since that makes it well suited for firefighters.
But a number of companies are working on inertial-only tracking, he says, and the ones he’s seen tend to fail. Sensors tend to drift, and tracking firefighters in particular can be difficult because in an emergency they don’t move as you would when walking down the street—they might go up and down staircases or crawl on their hands and knees. “I hope he really tests it out in those kind of conditions,” Duckwoth says.
Donikian says motionDNA takes drift into account. He says the physical device should be able to withstand fire, water, and major impacts and can alert others if you’re motionless for 30 seconds, showing where you are and whether you’re lying on your face or back. He hopes to start selling it for training purposes in the next few months at a cost of $4,000 to $5,000.
Beyond that, though, Donikian hopes to take Navisens to the mobile market, where it could augment existing technologies on phones. For example, it could correct errors that GPS may make outside in dense urban areas and help with Wi-Fi positioning systems indoors.
He plans to license the motionDNA technology later this year to software developers, who could use it in smartphone apps. A mall might use it with its own app to help shoppers find their way to specific items within stores, for example, or it could help pave the way for smarter buildings that turn lights and air conditioning on and off depending on where you are.
Duckworth can see the utility of consumer applications, and as he notes, many others do too. “Big players, small players—everyone’s trying to come up with the technology that’s going to enable it to happen,” he says. “Whoever’s going to crack that first is going to make a lot of money, I’m sure.”