My research focus is on developing a distributed platform for the execution of live streaming analysis. Consider, for example, the Internet of Things (IoT) that will very soon bring us massive amounts of data about the real world, provided live as continuous data streams from myriad sensors. The potential for amazing M2M and cyberphysical applications is staggering, from intelligent spaces to smart, connected vehicles. Currently, people working on such applications have to create turnkey solutions, that is they build the entire vertical from infrastructure up to application algorithms. Unfortunately, this is not conducive to the proliferation of such applications. The focus of my research is to create systems support for these applications, managing devices, infrastructure, computational resources, communication, etc. so that developers can focus on creating exciting new application software!
I have created a system that manages distributed, computational resources and facilitates executing IoT analysis applications that include rich algorithms, e.g., including computer vision and machine learning. It also provides support for the component-based development of such applications, as well as for sensor stream management.