In order to justify their name, “smart glasses” need to recognize objects and determine their location in the user’s field of view. While different technologies can be used for object recognition and position estimation, in mobile devices there is a critical tradeoff between accuracy and battery life. A camera, for example, can provide high accuracy, but with its high-power consumption it cannot be on for longer periods. Infrared technology requires an infrared (IR) emitter and receiver with relatively high power needs. Radio frequency (RF, as in RFID tags) has relatively low power needs, but difficulties with accuracy.

The authors propose a hybrid approach, combining RF and IR technology, where RF provides identification information and synchronizes very short IR emissions for position determination. In a demonstration prototype setup, they added IR emitters to RFID tags, and an array of IR photodiodes to a smartglass frame. This setup relies on a protocol where the transmission of an RFID packet triggers a very short, high-intensity IR burst. This keeps power consumption at the sender very low, and enables the receiver to turn on the IR receptors in time to measure the intensity and distribution of the incoming IR pulse, from which the position of the sender can be derived. With an accuracy of about 97 percent for recognizing objects, and a battery life of years for the tags and days for the receiver, the overall performance appears to be suitable for use in smart glasses. Collisions of signals in space (tags in close vicinity) and time (tags emitting signals at around the same time) are within acceptable ranges, especially when restricted to objects within the viewing angle of the user.

Of course this approach still relies on attaching tags to objects, so it is unlikely to revive broad use of smart glasses. But for tasks such as taking inventory in a warehouse, just wandering around and looking at the goods sounds more appealing than manually counting or scanning barcodes.