Learning to Compute without Reliable Power

The Internet of Things is nothing but hype until we learn to compute without batteries. We are simply not going to recharge, replace, and dispose of trillions of batteries. But, hope remains. Battery-less sensing devices are poised to transform science and society by enabling long-term maintenance-free data gathering, but we currently lack the hardware platforms, abstractions, languages, and tools that we need to harness this potential. These devices replace batteries with capacitors, which are smaller, cheaper, more environmentally friendly, and can work for decades (most batteries wear out after 2-5 years). The tradeoff is they don't store as much energy, and they fail more often. Even with energy harvesting advances, today's batteryless devices are difficult to program, test, and deploy, due to unpredictable energy supplies, limited energy storage, and frequent power failures.

In this talk, I will describe how we are learning to compute in the face of unreliable power. I will describe hardware and software, tools and techniques, and new programming models for building tiny sensing systems that depend on harvested energy, that can be deployed for long periods of time without battery changes, and that are able to adapt to uncertain energy conditions and thrive in spite of frequent power failures.
Jacob Sorber is an Associate Professor and Dean's Professor of Computer Science at Clemson University. His work makes mobile sensors and embedded systems more efficient, robust, deployable, and secure, by exploring novel systems (both hardware and software) and languages. His research is supported by the NSF (including a CAREER Award), the USGS, General Electric and other sources. He works on problems in health, biology, agriculture, and manufacturing and received the Best Paper Award at ACM SenSys 2014. Before joining Clemson, he was a postdoctoral researcher at Dartmouth College, a graduate student at UMass Amherst.