SensoryPhone takes a sensor rich mobile device as its starting point. While such a mobile devices possess a multitude of sensing capabilities in hardware, the actual software applications are highly restricted in using the sensors due to battery limitations. The SensoryPhone is a novel software and hardware architecture that enables extensive use of sensing resources in an energy efficient manner, to enable applications to maintain constant awareness of user context and inferred activity states.

Project components:

• A-Loc: The current location as well as the past location trajectory of a mobile device is immensely useful for several applications, both on the mobile device, such as searching for nearby services, and for external online services such as optimizing home heating controls based on user's expected arrival. However, continuously sensing location using GPS or WiFi/Cell-ID based online location service lookup will drain the mobile device battery very quickly. We develop energy efficient methods that intelligently determine when to sense location and which location modality to use depending on expected sensor error and application accuracy requirements, in order to create the appearance that the mobile device always knows its location.
• LittleRock: The current phone architecture is not designed for continuously running the phone's sensors since the processor energy required to access and process the sensor data is prohibitively high. We have developed a new phone architecture, LittleRock, that enables always-on sensing without noticeably impacting battery life. A low power processor is introduced into the system to continuously sense at power levels that are lower than the phone's standby power, and intelligent software mechanisms are used to activate the primary phone processor only when necessary to take a significant action upon the sensor data.
• Falcon: Falcon uses context information to improve OS performance for multiple OS services and application tasks. As a specific example, we show how the use of context, such as user location and temporal behavior, can be used by the OS to pre-load applications into the memory and provide rapid application launch. Falcon buils system support to introduce context based user behavior learning into the OS and allows specifying OS and application actions for various contextual states.[More...] [News]
• LEAP: GPS is one of the most energy hungry sensors. A large part of the energy is spent on processing the received radio signal in the GPS chip. The LEAP project introduces new methods for GPS signal processing that leverage additional information available on smartphones as well as cloud offloading to reduce the overall energy cost of using GPS based location for latency insenstive location tracking scenarios.
• SpeakerSense