Indoor navigation goes hybrid
Chicago O’Hare Airport has 17 different Starbucks. The line at the Starbucks nearest your departure gate is startlingly long – so what’s the quickest way to find another Starbucks? There’s an interactive map of O’Hare Airport, complete with all the Starbucks, but since you’re so caffeine-deprived, you’re having a difficult time reading it. Wouldn’t it be nice to have step-by-step directions to the two closest Starbucks?
Or what if you’re walking around a massive shopping mall and are looking for a particular store that you know is having a sale? Indoor navigation on your smartphone or your smartwatch would allow you to find a particular restaurant or store in real-time, relative to your current location.
Smartphone and wearable designers want to deliver more accurate indoor navigation to consumers, in large part because of demand from carriers and data aggregators (like Google) who will work to develop new revenue streams enabled by indoor navigation.
While some level of indoor navigation exists, at least in some places, current solutions leave much to be desired. Google Maps and GPS-enabled devices and smartphones have enjoyed tremendous adoption and are considered indispensable tools by consumers.
However, it has been challenging to deliver the same functionality indoors. GPS has historically been the most prominent positioning technology in the outdoor environment but it cannot provide adequate positioning indoors, with its weak signals unable to penetrate walls effectively. This is a major deficiency since mobile devices and smartphones are typically used inside rather than in outdoor locations.
Furthermore, there is a strong impetus to enable services that empower consumers while also providing commercial monetization opportunities. The ability to acquire accurate, granular indoor location data is poised to open up huge opportunities in a variety of markets such as proximity-based mobile advertising, augmented reality, retail, healthcare and public services.
Wi-Fi triangulation and Bluetooth beacons are existing technologies that are competing to enable indoor navigation. While several competing standards are deployed in a few showcase locations — a handful of airports, shopping malls and exhibition centers — Wi-Fi and Bluetooth beacons are difficult to roll out in a ubiquitous manner because they require:
* Infrastructure set-up with cooperation from venue owners, and timely updates about the location of each access point/beacon
* Handset Bluetooth and Wi-Fi that are always on, which rapidly drains power and inconveniences users
Each of these technologies also faces significant challenges in terms of accuracy. Specifically, Wi-Fi location technology is accurate to approximately 5 to 30 meters or more depending on Wi-Fi signal attenuation, which varies in the presence of people and objects and the location of the Wi-Fi access point. In order to achieve an approximate 5m level accuracy, the precise location of each access point must be known, and a fingerprint database developed.
Communication Service Providers (CSPs) are able to achieve at best 10 meters accuracy from LTE diverse location determination and delivery capabilities. In a perfect implementation, Nokia’s High Accuracy Indoor Positioning (HAIP) Bluetooth can be accurate to about 0.5m – 1m, but will require substantial modification to current Bluetooth 4.0 chips and significant investment in Bluetooth beacons.
While ~10 meter accuracy is sufficient for basic store-level location tracking, the market will ultimately demand sub-1m accuracy where one can identify if a consumer is at a specific position, such as in front of a particular display or aisle. Hence to date, rapid rollout of value-added indoor services has been inaccurate, delayed and spotty. Consequently, both new use cases and uptake by consumers has been slow.
Given the limitations of existing solutions, what type of approach could produce the highest-accuracy indoor navigation today?
Wi-Fi, cellular and pedestrian dead reckoning (PDR) using MEMS motion sensors in mobile devices (namely sensor fusion of gyroscopes, accelerometers and magnetic sensors) comprise a hybrid approach.
In fact, ABI Research projects that by 2014, hybrid solutions (with Wi-Fi, Bluetooth and sensor fusion) will have already surpassed standalone indoor location technologies on smartphones, with Wi-Fi and sensor fusion hybrid solutions reaching over 900 million units in 2018. Accurate, low-power sensor fusion (including key algorithms for PDR) are essential to implementing this hybrid indoor navigation approach.
George Hsu is president and CEO of PNI Sensor Corporation, which has developed the PDR portion of the hybrid solution with a sensor fusion technique that combines data from multiple sensors intelligently, correcting for the deficiencies of those individual sensors in order to track position accurately (down to 1m), either with or without the presence of infrastructure or other complementary positioning technologies. .
He has focused his 20+-year career on the sensor industry, having invented several magnetic sensor breakthroughs, including the Magneto-Inductive technology that is the core of today's electronic compassing in both the automotive and consumer markets. George is a graduate of Stanford University School of Engineering, holds several patents, and is a much-published author of technical articles on sensor theory, design and applications.