Trustworthy Digital Twins for Smart Buildings

A project using digital twins to develop statistical models with disclosure control for private information, as well as and the design GDPR-compliant digital twins.

In this project we study the efficient algorithmic design of trustworthy digital twins for smart buildings. The goal is to develop the design methodology that results in statistical models where the disclosure of private information is controlled with analytical guarantees and the model is representative for the application in smart buildings such as HVAC monitoring or advanced energy management.

Digital twins and GDPR

We will also explore how to design digital twins that are compliant with GDPRs data minimization principle. The research will build on previous research on privacy-preserving energy management design (see figures below, data collected in KTH Live-in Lab, [ACCESS, TIFS]), digital twins for HVAC systems, cross-disciplinary research on law and technology [FIU], and our recently proposed privacy measure called pointwise maximal leakage [PML]. 


[ACCESS] R. Reddy, T. J. Oechtering, and D. Månsson, “Adversarial Inference control in Cyber-Physical Systems: A Bayesian Approach with Application to Smart Meters,” in IEEE ACCESS 2024.

[TIFS] ] Y. You, Z. Li, and T. J. Oechtering, “Energy Management Strategy for Smart Meter Privacy and Cost Saving,” in IEEE Transactions Information Forensics & Security, vol. 16, pp. 1522-1537, 2021. 

[FIU] T. J. Oechtering, S. Saeidian, and C. Magnusson-Sjöberg, “Calculated Privacy: Tech Meets Law & Law Meets Tech,” in FIU Law Review, 17(2), Jan 2024.

[PML] S. Saeidian, G. Cervia, T. J. Oechtering, and M. Skoglund, “Pointwise Maximal Leakage,” in IEEE Transactions Information Theory, vol. 69, no. 12, pp. 8054–8080, Dec. 2023. 

Project contact

Tobias Oechtering

Tobias Oechtering

Division of Information Science and Engineering, School of Electrical Engineering and Computer Science