Skip to main content


This project entails developing a building cyber-security platform – BUILD-SOS via a probabilistic approach. BUILD-SOS protects smart buildings by effectively detecting faults and robustly operating the automation system, considering the uncertainties and probabilities in the control lifecycle. Four objectives are proposed to achieve the project goal:

  • Probabilistic graphical models for detection and diagnosis
  • Security constrained stochastic look-ahead optimization with vulnerability assessment and contingency analysis
  • Probabilistic post-control performance assessment
  • Validation through Building Cybersecurity Testbed (BCST) and real building demonstration

Nothing is certain, and uncertainties come from building modeling, data sources, and even the unobservability of adversarial events. The BUILD-SOS platform is based on advanced data analytics and stochastic optimization. The probabilistic approach greatly increases the attacking difficulties. Hackers need to deeply understand both probabilistic detection algorithms and stochastic controls in order to execute any effective attacks. Buildings are fully armed with the BUILD-SOS.


This project is well-aligned with BTO’s Grid-Interactive Efficient Buildings (GEB) program area. The resulted BUILD-SOS provides a holistic solution to secure building operations and can be broadly applied to commercial buildings and campuses that are prone to cyber threats. The project will transform the way buildings are currently operated, and greatly benefit our living and work environment.

Funding: DOE BENEFIT program

Principal Investigator

Qun Zhou Sun, Ph.D.
Associate Professor of Electrical and Computer Engineering


Zhihua Qu, Ph.D.
Thomas J. Riordan and Herbert C. Towle Chair and Pegasus Professor of Electrical and Computer Engineering
Wei Sun, Ph.D.
Associate Professor of Electrical and Computer Engineering