FuseForward and Toronto Metropolitan University Gain Better Data Visibility and Optimize Efficiency with Smart Campus Platform on AWS

Toronto Metropolitan University is a public research university located in the Garden District of downtown Toronto, Ontario, Canada, with approximately 50,000 students and staff on campus. For the past several years, the university has been outfitting buildings on campus with Internet of Things (IoT) sensors to collect data on everything from HVAC systems and energy and water meters to Wi-Fi nodes. For example, the university’s Daphne Cockwell Health Sciences Complex, a 332,604-square-foot building with 29 stories of classrooms, laboratories, offices, and student housing, has close to 14,000 data points collecting information every day.

University researchers wanted to bring all the data from the disparate IoT systems across buildings together to gain new insights. “These are siloed systems that were never designed to speak to a third-party integrator,” says Jenn McArthur, associate professor in the Department of Architectural Science at Toronto Metropolitan University. “We wanted to find a way to unlock these systems and actually get useful data.” The university also wanted to use its campus as a smaller model for a smart city. “We see the campus as a great test bed for smart city technologies,” McArthur continues. “We knew we had to have a partnership with someone to help us develop smart algorithms and scale things to demonstrate the power of smart buildings.”

To realize its smart building vision, the Toronto Metropolitan University Smart Building Analytics Living Lab started working with FuseForward, a Canadian provider of cloud and managed services and technology solutions. An Amazon Web Services (AWS) Partner, FuseForward specializes in AWS-based solutions that use IoT and other smart technologies. FuseForward was also the launch partner for AWS IoT TwinMaker, a service that developers can use to create digital twins, or replicas, of real-world systems such as buildings and factories. “We were interested in AWS, and FuseForward was a natural fit for us in terms of the work we wanted to do,” says McArthur. 

Toronto Metropolitan University became a member of the FuseForward Intelligent Systems Alliance, a network of academic and industry partners exploring advanced analytics, and also received a grant from the Natural Sciences and Engineering Research Council of Canada (NSERC), a federal funding agency for university-based research. “FuseForward is our industry partner in data services and modeling, and our NSERC grant supplements that work,” McArthur adds.

FuseForward helped Toronto Metropolitan University develop a comprehensive, integrated platform based on AWS services. “We created a smart campus platform for collecting data from smart buildings, smart transportation, and smart infrastructure applications,” says Michael Lamoureux, vice president of research and lab services at FuseForward. The platform relies on the Amazon OpenSearch Service and MySQL databases to run queries and analyze data from air quality monitors, energy meters, and temperature sensors in the Daphne Cockwell Health Sciences Complex. Toronto Metropolitan University analyzes live data with Amazon Kinesis Data Streams and AWS Lambda, a serverless compute service. Amazon Kinesis Data Firehose processes the data and stores it in Amazon Simple Storage Service (Amazon S3) buckets. Toronto Metropolitan University researchers use AWS Glue to prepare and process data and visualization dashboards running on Amazon Elastic Compute Cloud (Amazon EC2).

In addition, Toronto Metropolitan University worked with FuseForward to design a digital twin of the Daphne Cockwell Health Sciences Complex that serves as a 3D visual representation of structured data from the building. Through the digital twin, Toronto Metropolitan University building managers have a real-time picture of equipment and spaces in the building while they simulate scenarios such as the impact of shutting down utilities. FuseForward provides cloud analytics expertise and tools to manage and analyze the data coming in from the buildings and simulations.

Collaborating with FuseForward, Toronto Metropolitan University has integrated building data from disparate systems into a single solution. “We like the fact that all this different building data is flowing into one place for researchers to view,” Lamoureux explains. “Currently, we’re working with one building, but eventually—with a full smart campus—there will be 50 or more buildings.”

Using the smart building solution on AWS, the Daphne Cockwell Health Sciences Complex currently has more than 14,000 sensors that generate 500,000 data records per day—adding up to 150 million records per year. Since implementing Amazon Kinesis Data Streams, FuseForward and Toronto Metropolitan University have increased the data retention rate from around 10 percent to more than 100 percent. “Instead of having the data sitting on a server in each building, it’s all on AWS, which means it can scale to take on larger amounts of data going forward,” says Lamoureux.

Toronto Metropolitan University researchers can now use the digital twin to monitor campus buildings in real time and create new data visualizations as needed. They can view and analyze space management and building automation system data from the science complex. “There is a large diversity of space types we can see, including labs, classrooms, offices, and residences,” says McArthur. Toronto Metropolitan University is using an interactive data visualization solution that will enable researchers to view updated data trends for equipment or spaces. For example, researchers can look at data and cross-check it with reported issues. McArthur explains, “We can view equipment data and use it to determine the root cause of faults.”

McArthur’s team also wrote an algorithm that uncovered a mislabeled CO2 sensor, which was labeled as a temperature sensor. “Building automation systems usually can’t tell when sensors aren’t working, so that was something we only discovered through the data science in this solution,” says McArthur. “The solution is giving us visibility into things we couldn’t see before. It’s almost like having a full-time energy manager watching every single piece of equipment and flagging things that don’t look right or aren’t working.”

Toronto Metropolitan University has started using machine learning to identify energy usage patterns and detect anomalies. “We would like to eventually predict when a boiler needs maintenance, which we could then proactively repair to prevent damage and save thousands of dollars,” says McArthur. “Using the FuseForward solution on AWS, we’re moving from a reactive to a proactive stance, creating data visualizations that help us predict building behavior and run the buildings more efficiently.”

The university’s department of civil engineering is using the digital twin campus to focus on 3D modeling and geospatial data analytics. “We are looking at thermal data analytics to try to monitor the energy efficiency for different buildings on campus,” says Ashraf Elshorbagy, a postdoctoral research fellow in the Department of Civil Engineering at Toronto Metropolitan University. For instance, Elshorbagy is using data modeling to assess the quality of insulation and see if there are any leaks. “Energy for heating is one of the largest expenses on campus, and we’re trying to get a holistic picture of all the buildings to help optimize energy efficiency,” says Elshorbagy. Toronto Metropolitan University plans to use this information to reduce energy costs. “If we can use the data to predict peak energy usage, we can minimize energy use and keep it operating within an ideal efficiency zone,” says McArthur. Achieving that goal would help Toronto Metropolitan University align itself with two of the United Nations Sustainable Development Goals, including ensuring access to affordable, reliable, sustainable, and modern energy and making cities and human settlements inclusive, safe, resilient, and sustainable.

Eventually, Toronto Metropolitan University will use the digital twin as a small-scale model of a smart city. “We anticipate integrating building data with infrastructure data like water, foot traffic, and electricity,” explains McArthur. “This will help us make more informed decisions about the entire campus.” In addition to continuing its use of the digital twin, FuseForward plans to scale up from its Toronto Metropolitan University project and continue implementing smart city technologies. Lamoureux concludes, “We will take our experience from the AWS Partner Smart Cities Pilot Program and work closely with city governments to develop smart city solutions for them.”