Tracking a Smartphone User Around the World

Organization: Princeton University

Patent: Pending

Reference: http://puotl.technologypublisher.com/technology/25745

Inventors: Arsalan Mosenia, Xiaoliang Dai, Prateek Mittal and Niraj K. Jha

Background:

GPS signals are not always available; being indoors, deep in foliage, or in a thunderstorm can block GPS signals and, therefore, interfere with tracking. In addition, some remote locations do not receive GPS signals

Invention Overview:

Novel user/cargo-location algorithm, named PinMe. This mechanism utilizes non-sensory/sensory data stored on a smart device (e.g. the environment’s air pressure and device’s time zone), along with publicly available auxiliary information (e.g. elevation maps), to accurately estimate a subject's location when location services (e.g. GPS) are turned off. While each such signal may not be highly discriminative by itself, the collection of such signals, along with robust machine learning, can provide location accuracy similar to traditional GPS services

Stage:

Proof of Concept

Applications:

  • Cheap and reliable method of navigation for self-driving cars and trucks
  • Forecast the most efficient/optimal supplier delivery. Provide efficient delivery tracking with audit trail
  • Track consumers within a store to optimize the consumer experience

Differentiators:

Unlike competitors in this space, PinMe neither requires any prior knowledge about the user, nor a training dataset on specific routes. The algorithm can discriminate between four activities (walking, traveling on a train, driving, and traveling on a plane) and accurately estimates a user’s location during each activity

Invention Disadvantages and Possible Mitigation:

Predictions/estimates of the position of an object uses the last known position (when it was connected to a WiFi network) and with the data from sensors and additional “public available data” (such as activity data -> drive, travel …) determines the new position. Combine with additional sensors to gain more precision on where the object is