Multilevel Cybersecurity for Photovoltaic Systems
Funded by DOE SETO program (UGA PI, $3.5M, 4/2020-4/2023)
The project goal is to devise a multilevel cybersecurity solution to address PV security gaps at the inverter and system levels, and field test the solution under the supervision and review of a US-based solar inverter manufacturer (GE) and PV installer/operator (TPI). The primary objectives are to:
- Create an inverter framework within which multiple new security capabilities can be deployed. These include supply chain security, hot patching, firmware protection, safety protocols, side-channel monitoring, controller security and control resilience.
- Deliver a self-consistent set of complementary system security methods that will work with data routinely available in PV solar farms.
Data-Driven Monitoring and Diagnosis of Industrial Machines via Electrical Waveform Auditing
Funded by DOD Airforce program, (UGA PI, $900K, 5/2020 – 5/2023)
The goal of this project is to design an Electrical Waveform Fault and Attack Detection System (EWFADS). The goal of the system is to monitor the condition of manufacturing machines to detect faults and cyberattacks by analyzing the electrical waveforms of the electrical power connection to the machine at the point of common coupling. In Phase II the team will perform additional research and development to:
- Refine and expand comprehensive mathematics models; case studies for physical fault and cyber-attack scenarios.
- Refine features and machine learning (ML) algorithms for detection and diagnosis.
- Design and construct Hardware-In-The-Loop (HIL) real-time testbed and hardware Testbed.
- Testing, evaluation, and improvement of the developed detection and diagnosis algorithms.
Collaborative Research: SaTC: CORE: Medium: Cyber-threat Detection and Diagnosis in Multistage Manufacturing Systems through Cyber and Physical Data Analytics
Funded by NSF SATC program (Co-PI, 1.2M, 10/2020-10/2024)
In modern multistage manufacturing systems, with increased software-defined automation and control as well as monitoring of manufacturing assets across networks, exposure to cyber-attacks also grows. The cyber-threats may compromise the integrity of manufacturing assets (manufacturing systems and processes, machine tools, fabricated parts), reduce manufacturing productivity, and increase costs. Some cyber-threats including integrity attacks are only partially observable in cyberspace alone, and therefore need to be detected and diagnosed through inter-dependency analysis of both cyber and physical signals. Thus, there is a significant opportunity in exploring physical signals, together with cyber signals, to advance trustworthy manufacturing system research and design.
This project brings together leading researchers from manufacturing systems, computer security, and electrical drives to develop and demonstrate a new methodology and tool for cyber-threat detection and diagnosis in multistage manufacturing systems. The cyber-security tool will monitor a variety of cyber and physical signals and perform cyber-threat detection and root cause diagnosis through advanced cyber-physical data fusion and taint analysis. The goal is to enable the prevention and mitigation of potential harms at the early stage, proactive and predictive maintenance, and countermeasures. This project attempts to integrate and analyze the process and quality signals, and the signals from the power networks and cyber networks of multistage manufacturing systems to detect and diagnose cyber-threats. This new systematic approach expects to capture new cyber-threats, especially data integrity attacks, that traditional cyber-security approaches may not capture. The proposed data analytics and methodology for integrating cyber and physical signals will advance a fundamental understanding of cyber-threat detection and diagnosis in multistage manufacturing systems and can broadly apply to other cyber-physical systems.