In 2020, as the COVID-19 pandemic was sweeping the globe, many local school districts were struggling with the decision on how schools should open for the 2020-2021 academic year. Was it safe to bring students back to the classroom? Should learning take place remotely? Should schools offer a hybrid educational experience, with time split between classroom and remote learning? For educational leaders wrestling with these questions, local data to inform decisions for their school district was hard to come by.
It was concerns like these that prompted the rise of an ambitious project based at The Ohio State University to provide educational administrators with the timely local information they need to help ensure a safe learning experience for students, staff and teachers. This initiative, known as the COVID-19 Analytics and Targeted Surveillance System, or CATS, puts data analytics and visualization tools to work to allow school superintendents and local public health departments in Central Ohio to make critical health and safety decisions with the confidence that comes with timely local data. CATS served the school-based monitoring needs of 21 local school districts in Central Ohio, serving approximately 1.4 million residents and 238,000 school-aged children during the COVID-19 pandemic.
The CATS project was spearheaded by Dr. Ayaz Hyder, an assistant professor in the College of Public Health at The Ohio State University and a core faculty member at the Translational Data Analytics Institute. The initial impetus came when he learned there was no clear strategy to help local school districts make decisions about when to switch between different learning modalities.
Dr. Hyder reached out to a local school district, where three of his children are enrolled, and proposed a pilot project using the COVID-19 monitoring methods he had developed for agencies throughout the state. The district agreed, and the local public health departments were brought on board to provide expertise in contact tracing, outbreak investigation and data interpretation. In the weeks and months that followed, additional schools and stakeholders joined the effort, as did a contingent of University student workers who assisted in programming, managing and visualizing the CATS data.
“The school districts were getting pressure from parents to use local data, because county-level trends don’t always represent what’s happening in the individual community,” Dr. Hyder says. “We were able to take that pressure off of the superintendents and school administrators by providing local data, in partnership with the local health departments and their epidemiologists and nursing staff. We could then make a strong case to parents and staff that local data was being used for local decision-making.”
How it works
The CATS application considers multiple data flows to help school leaders and public health officials make informed decisions on the most appropriate learning modalities — virtual, in-person or hybrid — to help prevent and control the spread of disease. These include the monitoring and epidemiological review of school nurse visits among students and absences among students and staff due to COVID-19-like illness. The team also considers temporal and spatial patterns in county-level and school district attendance area-level data for COVID-19 case rates.
The information from these diverse data flows is entered CATS. Using password-protected CATS dashboards, school district staff and the local public health department monitor the system, watching for signs of coronavirus outbreaks. The general public is also able to access CATS dashboards that contain aggregated information specific to each school district. By closely tracking possible indicators of COVID-19 infection, the CATS system allows for real-time analyses of factors that can help predict localized outbreaks.
In addition, CATS includes web-based applications, automated alerts, and weekly reports for the general public and decision makers, including school administrators, school boards and local health departments. The result is essentially an early-warning system for case clusters — and a safer educational experience for students and their teachers.
Drawing on the expertise of the Ohio Supercomputer Center
The CATS project was a computationally-intensive undertaking that brought its own set of technical challenges. To overcome these challenges, the CATS team turned to the Ohio Supercomputer Center in Columbus, Ohio. OSC is dedicated to making high performance computing resources and expertise readily available to university and industrial researchers in Ohio and points beyond.
In these efforts, the CATS project benefited from the computational power of the HPC systems at OSC made accessible via Open OnDemand. This portal, developed at OSC, gives users access to HPC systems through a web browser. That made it easy for the CATS team, including its software developers, to run HPC workloads.
The systems that are accessible via the Open OnDemand portal include OSC’s two main clusters, Owens and Pitzer. Collectively, these clusters built by Dell Technologies deliver the power of more than 50,000 Intel® Xeon® compute cores, along with hundreds of GPUs.
“With the resources of the Ohio Supercomputer Center, we were able to set up 16 different dashboards that were going to be accessible by hundreds and thousands of people at the same time,” Dr. Hyder says. “We were really fortunate to have that kind of support from the Ohio Supercomputer Center.”
For the full story, see the Dell Technologies case study “Innovating with data for the
safety of children” and the video “Schools work with Ohio Supercomputer Center (OSC) for safety.”