Project Title

Smart analytics for smart grid


Researcher

Andriy Miranskyy, Ryerson University


Industry partner

IBM Canada Ltd.


Supported by

SOSCIP, IBM Canada Ltd., OCE, FedDev, Ryerson University’s Faculty of Science

Cybersecurity Energy

How we use energy reveals much about everyday behaviour – from our work and school schedules to when we are likely to be sleeping, eating and performing simple chores such as laundry or washing dishes. It even reveals when we are on vacation.

While this information in the wrong hands could put us at risk, it also plays an important role in reducing our carbon footprint and conserving energy.

Power generation companies use information derived from micro­climate, small areas within general climate zones, to accurately predict power generation needs. Microclimate information is obtained by analyzing energy consumption data sent by the smart meters from our homes to utility companies. 

With more than one million smart meters installed in Ontario homes since 2010, it’s necessary to find ways to ensure the data captured by smart meters is protected and secured.

Prof. Andriy Miranskyy from Ryerson University’s department of computer science, and his team, are conducting research to help power generation companies capture the data they need to accurately make these predictions while also respecting consumers’ privacy.

The team includes IBM Canada’s Dr. Biruk Habtemariam, Prof. Ali Miri from Ryerson University, Prof. Saeed Samet from Memorial University and Prof. Matt Davison from Western University. Their goal is to develop data aggregation and obfuscation techniques that will prevent privacy violation, yet yield the information needed for accurate prediction of power consumption. 

“This will enable utility companies to open up data to external groups without jeopardizing the privacy of the millions of people they’re collecting data from,” explained Prof. Miranskyy.

Through SOSCIP’s Cloud Platform, they can develop and test an algorithm using simulated data which will allow them to predict electricity usage in near-real time by a third party while keeping the data secure. 

“Essentially, the secure linear regression algorithm allows researchers to compute numbers without actually being able to view those numbers. This allows utility companies and other third parties to provide better services to Ontario residents, leading to energy conservation and reduced infrastructure costs.”

The team’s work, Privacy Preserving Predictive Analytics with Smart Meters,
has been published in the Proceedings of the 5th IEEE International Congress on Big Data, 2016.