escience2019

Scientific instruments have changed dramatically over the last century. We have moved from hand-crafted telescopes and thermometers to advanced digital sensors that can continuously provide data from the planet’s most remote locations. Recent advances in machine learning technology are leading to another significant shift in sensor science and scientific instrumentation — software-defined sensors. By embedding parallel computation with high-bandwidth sensors, a new kind of intelligent, software-defined edge computing sensor is possible. This disruptive new approach to scientific observation is part of an evolving computing continuum that analyses data in-situ, and uses HPC to model, predict, and learn. In this presentation, the Waggle Platform for building distributed edge sensors, developed at Argonne National Laboratory, and the Array of Things project at the University of Chicago will be discussed in the context of edge computing, machine learning, and the computing continuum.