CNS Core: Small: Low-Power Wide-Area Networks for Industrial Automation

About This Grant

The evolution of Internet of Things (IoT) is transforming the field of industrial automation including process control and smart manufacturing into an important class of Industrial IoT (IIoT). Today, wireless solutions for industrial automation are based on short-range wireless technologies (e.g., WirelessHART, ISA100). To cover a large area with numerous devices, they form multi-hop mesh networks at the expense of energy, cost, and complexity, posing a big challenge to support the scale and wide-area of today’s IIoT. For example, the East Texas oil-field extends over 74x8 square kilometers requiring tens of thousands of sensors for automated management. Also, in process industries, many silos, tanks, and plants are often positioned far from the center, at inconvenient locations in difficult terrain or offshore. Pipelines can be hundreds of miles long and pass through difficult terrains, making it difficult to monitor gas and chemical leaks in real-time. This project proposes to adopt the Low-Power Wide-Area Network (LPWAN) technologies for industrial automation. Due to long-range, LPWANs can be adopted without complex configuration and at a fraction of costs for wide-area IIoT applications, compared to multi-hop solutions. This project will develop theoretical foundations and systems for enabling industrial automation using LPWANs. Its important findings will be shared with the standards bodies and industries. The developed technologies will be made open-source. This project will particularly consider LoRa, a leading LPWAN technology. Adopting LoRa for industrial automation poses some evolutionary challenges. The fundamental building blocks of any industrial automation system are feedback control loops that largely rely on real-time communication. Due to severe energy-constraints, LoRa uses a simple media access control protocol that is unsuited for real-time communication. It needs to adopt low duty-cycling in several regions (e.g., Europe). In addition, to optimize performance, industrial automation needs a codesign of real-time scheduling and control. Such a codesign becomes specially challenging in LoRa because it is large-scale and has energy-limitations. This project will address these challenges and make the following contributions: (1) an autonomous real-time scheduling technique and analysis using the demand bound function theory for LoRa; (2) a scalable scheduling-control codesign that jointly and dynamically determines control input and sampling rates; (3) a highly energy-efficient codesign by maximizing the sleeping times of the devices through a combination of self-triggered and even-triggered control adopting state-aware communication; and (4) an evaluation of the results through experiments using industrial process control use-cases on a physical testbed. 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.