With the release of the new iPhone from Apple earlier this year and the ability of its operating system (iOS 4.0) to more effectively run background apps and services, there are a host of possibilities for new location-based services. And as anyone developing apps on the Android platform will happily tell you, its location framework has supported this capability through the LocationManager system service for years.
One emerging technology that will likely play a large role in driving enhanced location-based services (LBS) is something called a geofence. In many ways, a geofence is simply another name for something that has been around for a long time - a geographic boundary (i.e. polygon) defined by a series of lat/long coordinates. The term geofencing refers to the use of geofences in combination with a location-aware device running apps/services that can send notifications when the person carrying the device crosses a geofence. The spatial processing (e.g. proximity calculations and probability estimates) needed to support geofencing applications is already in use for a host of other LBS.
There are several types of geofences and a seemingly unlimited number of ways to apply them to mobile applications. Location Labs, a mobile location app provider, has defined static geofences as those used in applications where a user’s movement is correlated to a static location. In static geofencing scenarios, notifications (e.g. instant messages or emails) are initiated based on a mobile user’s relationship to a place or predefined boundary. In the case of place-based static geofences, a radius perimeter is typically used to define the boundary that triggers events. In the case of predefined static geofences, the delimiter used to initiate actions or notifications is based on a known geographic area (e.g. neighborhood).
While many LBS app developers will use radius-based geofences because they make sense for the service they wish to offer or because they are easy to generate, there are a range of services that are only possible by leveraging predefined geofences. There are a host of predefined boundaries that may be relevant in geofencing applications, including everything from custom polygons that represent established delivery territories to neighborhoods. Below are a few use cases that argue for using common, predefined areas (neighborhoods, in these examples) as the best perimeters to drive user alerts in geofencing applications.
Predefined Geofences vs. Radius-based Geofences
Many common geofencing scenarios are based on a simple radius around a point of interest, like a retail location. A radius geofence may be used as part of push-marketing campaigns where store managers hope to bring in nearby shoppers by offering promotions and discounts. While users opt-in for such services and are presumably seeking these kinds of location-based ads, the fact that notifications may be sent when users cross any number of radius-based geofences in an area can easily overload them with options.
Contrast this with predefined geofences, which offer a broader way to define a location. As a user, you could set your preferences to notify you when there are restaurants offering specials in the neighborhood where you and your friends are hanging out for the evening. Instead of a slew of individual text messages, you could visit a single mobile app and check the list of offers and of course, get exact directions to take advantage of them. This type of mobile ad model may be more sensible and user-friendly in the long-run and it requires predefined geofences for neighborhoods. Neighborhood boundaries are meaningful to users specifically because they are defined by the people who visit and live in them.
A sense of location and mobility is becoming a fundamental part of social networking. Thus far, much of the emphasis has been on tagging individual posts with location information (read about how Maponics powers place designation for geotagged Tweets). But generating alerts and Tweets based on when you and your friends enter or leave areas like choice neighborhoods can give you a heads up about where the party is headed. Also, using neighborhood boundaries as geofences enables alerts based on when other friends enter the neighborhood where you are hanging out. In these types of use cases, social networking users can choose to share some location information but stop short of divulging exact location - something that many users clearly worry about.
Mobile apps and house hunting go well together. Consider the following scenario:
A couple with two kids in elementary school is planning to relocate to a new city for career opportunities. Prior to their house hunting trip, they visit an online real estate website and generate a short list of properties to visit based on numerous preferences, including price range, number of bedrooms/bathrooms, amenities, etc., within several neighborhoods that they’ve heard have good schools and are kid-friendly. Further, on the real estate website, they sign up for a geofencing service to be notified on their smartphone (the house hunters tool of choice) when they enter and exit their target neighborhoods so they can get a sense of locally-recognized neighborhood boundaries and to help orient and direct their house hunting activities.
The couple uses the online real estate site’s mobile app on their iPhone to get directions to the first few properties. While looking at one of the houses, they visit a nearby park with their kids and strike up a conversation with a local mom. They quickly find out that there are several public elementary schools nearby but they are not at all on the same level as far as quality and ranking - and that the attendance areas for the schools cut across several of their target neighborhoods.
Armed with this local "intel" they launch the online real estate app once again and find, to their surprise, that area schools, GreatSchools ratings and even school attendance zones are search parameters they can use to further refine their target house list. Similar to the neighborhood alerts they set earlier, they turn on the geofencing service for elementary school attendance zones so that as they drive around they know when they enter and exit the exact boundaries that define which public schools residents attend.
The couple’s location-aware, micro-targeting saves them hours of wasted time looking at houses that on the surface may look fine but ultimately will not work because of the location in relation to desired neighborhoods and schools.
Many parents grapple with striking the right balance when considering technology that can help them monitor their kids’ locations and activities. Lots of kids now have GPS-enabled phones, so it is easier than ever to track them, but exact location tracking is not always a desirable option because kids reject it as too intrusive. With neighborhood-based geofencing, parents can get notifications if and when their kids enter certain neighborhoods - without monitoring their exact location. This might be just enough information for parents to know if they need to touch base or take action, but not so much as to be considered draconian.
Putting Geofences to Use
It is clear that companies will dream up all sorts of uses for geofencing but unless they respect users’ privacy and give them clear and easy ways to control how, when and with whom to share their location, adoption will be limited. The key to more wide-spread adoption will be striking the right balance between granularity and simplicity when designing and presenting location-sharing options.