CAREER: Enabling Dynamic, Adaptive, and Reliable Battery-free Embedded Computing

Northwestern University, Evanston IL

Investigators

Abstract

For decades, embedded computing and sensing systems have relied primarily on battery power. Yet, batteries are bulky, expensive, high-maintenance, and not sustainable for the next trillion devices. Instead of relying on energy stored in a battery, an emerging class of computing devices harvests all energy needed for operation from sources such as the sun, motion, radio waves, and vibration. However, building sophisticated applications on these battery-free systems is challenging due to frequent power failures from fluctuations in energy harvesting. Programmers must figure out how to string together fragments of execution to meet application goals while dealing with novel software and hardware bugs that stem from power failures. Because of this, memory-intensive, inference-heavy, and user-facing applications have rarely materialized on battery-free devices. New general-purpose hardware platforms with accelerators and heterogeneous computing modules are needed to build these applications. However, hardware is not enough. With new hardware comes new challenges like scalability, dynamism, and memory-efficient checkpointing. This project explores intermittent computing systems and toolchain support for integrating diverse computing modules, like FPGAs, Accelerators, and Vector Processors, alongside traditional microcontrollers. The project weaves scalability across the intermittent computing system stack, leveraging these new modules to enable reactive, adaptive, and high-performance applications on this important new class of computing devices. This project will explore and prototype scalable hardware platforms, adaptive software systems, high-level programming languages, and energy introspection tools that enable even novice developers to quickly prototype sophisticated battery-free applications, despite power failures. These advancements will be demonstrated and evaluated in the context of real-world deployments in mobile health, habitat monitoring, and interactive devices. Battery-free embedded systems offer a transformative and ecologically sustainable approach for building the next trillion computing devices. This project fills a gap for system designers who lack the hardware platforms, efficient runtime systems, and focused tools to build capable, data-intensive, reactive, and reliable applications with these devices. The results of this research will impact fields across scientific and industrial interests: including healthcare (wearable and body sensor networks), ecology, horticulture, infrastructure, conservation, and public utility monitoring, and many other areas where long-term, massive scale sensing is essential. The hardware, systems, and tools will speed up research and commercialization in critical sectors like smart cities and the Internet of Things. The project's demonstration applications, including smart health devices, interactive devices, and novice-focused programming environments, will provide proof of approach to encourage uptake of battery-free devices. The project includes outreach and education initiatives focused on increasing participation among Native Hawaiian youth in computing by introducing computing concepts via building sustainable and conservation-focused embedded systems applications in partnership with a Native Hawaiian serving public school and non-profit organizations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.