Smartphone-based location sensing for low-power vehicular navigation - Embedded.com

Smartphone-based location sensing for low-power vehicular navigation

This paper describes eNav, a smartphone-based GPS navigation system with a novel power-conserving mode. The system makes a contribution to low-power location sensing in the context of vehicular navigation. The distinguishing feature of location sensing in the context of vehicular navigation is that the location estimate does not have to be accurate at all times for navigation errors to be prevented.

Rather, it is allowed to get inaccurate in certain situations (e.g., when the vehicle is far away from the next navigation waypoint, while needing to remain accurate in others (e.g., when a waypoint is near). Hence, energy can be saved via an adaptive mechanism that keeps the location estimation error below a bound that changes depending on the current situation.

The mechanism judiciously switches between a cheap inaccurate estimation mode and an expensive accurate one, depending on the allowed location estimation error at the current time.

The motivation for this paper comes from the observation that smartphones have become popular means for navigation in vehicles. Dedicated GPS navigation devices, such as Garmin, see a continued decline in market share, whereas integrated dashboard systems are still an expensive option, compared to smartphone applications. Unfortunately, the GPS module is one of the most power-hungry components on phones

It may deplete batteries within hours (or less when the phone is not fully charged), running the risk of navigation loss while driving. The above observations beg the question: would an energy-saving mode be a useful addition to current phone-based GPS navigation applications used by drivers? If so, how should it be implemented?

In this paper, we first show results of a user-study that an wers the first question in the affirmative. We then present the design, implementation and evaluation of such a service, demonstrating significant energy savings.

Briefly, eNav allows the user to enter or exit an energy saving navigation mode at will. In that mode, two mechanisms are employed that reduce energy consumption; adaptive GPS sampling and screen saving. Adaptive GPS sampling refers to substituting actual GPS positioning with dead reckoning using cheaper sensors, whenever such a substitution is deemed safe.

The substitution is deemed safe as long as it cannot lead to a navigation error (for example, it is safe when the vehicle is sufficiently far from the next navigation waypoint ). Screen saving refers to turning the screen off, ostensibly to save energy, but in reality to mask the fact that location estimation is inaccurate at certain parts of the route.

As a waypoint approaches, the allowable location estimation error shrinks, GPS sampling restarts, and voice navigation alerts the driver to needed actions, making it look as if location estimation was accurate all along.

Both of the above mechanisms contribute to improved energy-efficiency, while keeping location estimation inaccuracy transparent to the driver. Importantly, the evaluation shows that adaptive GPS sampling  significantly increases savings over screen saving alone (which we use as a baseline for comparison).

According to a survey we conducted of 500 drivers, more than 37% said they ran out of battery while using a phone for navigation, and as much as 91% said they would like to have a vehicular navigation application with an energy saving mode. A user test-study of eNav shows that it reduces navigation energy consumption by around 80% without compromising navigation quality and user experience.

*** The other authors of this paper are Lu Su,State University of New York at Buffalo; Hengchang Liu,University of Science and Technology of China; Suman Nath,. Microsoft Research Redmond; and Romit Roy Choudhury and Tarek F. Abdelzah, University of Illinois at Urbana-Champaign.

To read this external content in full, download the complete paper from the author online archives at the University of Illinois.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.