While location-based services are bringing a lot of context and relevancy to the world, the technology is a lot harder to use indoors. But a Finnish company called Walkbase is looking to bring to market a very simple and scalable solution for indoor positioning that can help map out rooms using crowd-sourced data and smart self-learning algorithms.
The company — which opened its API to developers in October and will soon be deployed in some upcoming third-party apps — is moving ahead with its own application to showcase how the technology works. It’s introducing a new app for Android called Checked, an automatic Foursquare check-in app that lets people check-in to locations without having to open their app. There have been previous auto check-in apps but they often wear down the battery and are not that accurate.
But Walkbase has a different approach. Through the network of apps that will use its API, it’s building up a database of room fingerprints every time someone uses its technology in an app. Each time someone checks-in, Walkbase scans the room for Wi-Fi and other signals and uses the data to define a room. The self-learning algorithm can filter out less reliable information and can adjust when certain Wi-Fi access points go down or move.
Checked users can then save their favorite locations on the app and the Walkbase technology can check them in when they arrive. The check-in is verified against Walkbase’s database, which can cut down on fraud and inaccuracy. If the database doesn’t have data on that location, the person can submit a new location entry to update the database. The database already has hundreds of thousands of check-ins thanks to a few apps that are already testing the technology. About 200 developers have signed up for Walkbase’s technology. Walkbase can not only define rooms but also determine which floor a user is on based on their surrounding environment.
Walkbase CEO Tuomas Wuoti said Walkbase can get to accuracy of three to five meters. But unlike other indoor positioning efforts, like Google’s (s goog) new indoor positioning on Android, it doesn’t try to pinpoint a user’s location on a map. It settles for a little less specificity and instead works to map out rooms. That allows the system to scale more easily and it also means it doesn’t require as much battery use because it’s not scanning at such frequent intervals. Other systems can run down batteries or require venue owners to submit maps of their malls or buildings.
“We’re relying on crowdsourced and noisy data,” Wuoti said. “We do a good job of making noisy data clear enough to do good positioning for commercial benefit. That’s where the biggest commercial value is when a user or customer walks into a room or a store.”
Wuoti believes that merchants and restaurant owner can use this technology to push relevant content to owners based on their location inside. So a restaurant could deliver its menu to a user who has checked in. Or a retailer could push out very specific advertising when a person walks into an area within store. The check-in data could also provide useful analytics about customer flow in a location.
Walkbase was created out of an Åbo Akademi University project in 2009 and recently got $500,000 in seed funding. Wuoti said he’s talking to a lot of bigger technology companies and developers and said the first apps leveraging its technology will appear by summer. The technology is limited to Android right now and won’t work on iOS because Apple (s aapl) hasn’t opened up its APIs.
It will be interesting to see how successful Walkbase can be. A lot of big players like Nokia (s nok), Microsoft (s msft), Google, Motorola (s mmi) and others are also working on indoor location technology. This is a big opportunity to get more targeted data on users and also fulfill user intent on what they might be looking to do. And it gives a way to better push out advertising, which could be annoying for some users but very appealing for retailers and merchants. Walkbase will have to get over the initial barrier in any crowd-sourced project, building up the initial database. But it could be an attractive way for developers to leverage location indoors. And at the very least, it could be another way for users to handle automatic check-ins.