A lot of work is underway trying to add location awareness to smartphones and electronics while they’re indoors. When outdoors, they can use GPS to track their own locations to within a few meters. But indoors, where there is no reception of GPS signals, location positioning is only recently becoming available.
Technologies for indoor localization fall into 2 categories: those that require that dedicated infrastructure be installed and those that do not require infrastructure.
Systems with dedicated infrastructure require that dedicated appliances, called Locators or Beacons, be installed in any site where localization will be provided. These locators track the locations of smartphones or other electronic devices as they move around the site, or, alternatively, transmit signals that enable smartphones and electronic devices to track their own locations. This is the approach taken by the majority of indoor location companies in the market today, including Apple’s iBeacon initiative and other technologies using Bluetooth or ultra-wideband beacons.
By contrast, systems that do not require dedicated infrastructure use sensors in the device combined with knowledge of the environment to track locations. This may include things like Wi-fi hotspots that already exist in the site, but will not require that any dedicated Locators or beacons be installed. Systems of this sort are called infrastructure-free.
Infrastructure free systems are generally easier to install, since they do not require that any infrastructure be installed, but they have a much bigger challenge in localization. How can a system with no dedicated infrastructure perform location positioning anywhere near as well as a system with dedicated infrastructure?
One approach to infrastructure-free localization that achieves high accuracy is to use a device’s camera to sense the surroundings instead of radio. In these visual approaches, a device’s camera is used to literally “watch” the environment as it moves, and track the device’s location by analyzing the images that are seen from the camera. In some cases the images are used to track movement, often doing so much more accurately than using in-device sensors. In other cases the images are compared to a database of images in a site, and then processed to determine the distance and orientation of the device relative to the image seen.
One company with visual infrastructure-free indoor localization technology already in the market is Accuware, a spin-off from long-time location industry player Navizon. Accuware’s technology uses a video stream from a device camera to deliver sub-meter accuracy localization of the device. In addition to smartphones and tablets, their technology is made to work on robots, drones, autonomous vehicles and other location-aware electronics. In this way their technology has much broader application than simply the mobile industry.
In a recent demonstration, the robot shown here was operating on a golf course putting green, simulated here at a conference. A video feed was sent from the robot to the Accuware system, which interpreted the significant points on the image and generated the location positioning shown on the map. In this demonstration, the robot was programmed with a “geo-fence” that would prevent it from moving onto the green, even if its controller tried to move it there, based on the real-time localization that Accuware generated from the video feed. The fact that the robot was able to stop right at the edge of the green (not shown) indicates accuracy of closer than a meter. In other demonstrations, Accuware shows their technology tracking a smartphone, a forklift in a factory, autonomous robots in a retail setting, and more.
Another approach to infrastructure-free localization is motion sensing, based on sensors built into the device (the same sensors that detect whether you are holding your phone vertically or horizontally), along with occasional other signals for calibration. Many systems of this sort use a device’s built-in accelerometer, magnetometer and gyroscope to measure step-by-step movement as a device moves around a site. If the starting location is known, motion sensing of this sort can maintain a location estimate. As inaccuracies compound, such systems calibrate themselves based on magnetic field readings or radio signals from already-installed Wi-Fi hotspots. These systems are generally achieving accuracies of 1–2 meters, but not sub-meter. GipsTech, Navin, Situm, Navenio and Infuse Location are just a few of these companies.
For a lot more detail on infrastructure-free localization technologies than given in this article, see this report on infrastructure-free indoor location technologies.
This is only a small part of the overall indoor location market, which comprises over 200 start-up companies as well as many major mobile and chip manufacturers. The infrastructure-based segment has tended to get more attention, due to some early successes with Bluetooth beacons and also due to Apple’s iBeacons initiative. But with infrastructure-free technologies improving and reaching more and more markets, the next indoor location applications you read about are likely to involve absolutely no infrastructure.