Carat: Collaborative Energy Debugging

Google Tech Talk September 27, 2012 Presented by Adam Oliner. ABSTRACT We aim to detect and diagnose code misbehavior that wastes energy, which we call energy bugs. This talk describes a method for performing such diagnosis on mobile devices and an implementation, called Carat, for iOS and Android. Carat takes a collaborative, black-box approach. A non-invasive client app sends intermittent, coarse-grained measurements to a server, which identifies correlations between higher expected energy use and client properties like the running apps, device model, and operating system. Carat successfully detected all energy bugs in a controlled experiment and, during a deployment to a community of more than a quarter of a million users, detected (and sometimes diagnosed) thousands of instances of buggy apps running in the wild. About the speaker: Adam Oliner Adam is a postdoc in the EECS Department at UC Berkeley, working in the AMP Lab. Before coming to Berkeley, he earned PhD in computer science from Stanford University, where he was a DOE High Performance Computer Science Fellow and Honorary Stanford Graduate Fellow. Adam received a MEng in EECS from MIT, where he also earned undergraduate degrees in computer science and mathematics. His research focuses on understanding complex systems.