Carnegie Mellon University
April 22, 2020

CMU Dashboard Will Help Inform State Decision-Makers During Pandemic

Model will compile data to better understand areas' public health, economic status

Jason Maderer
  • Marketing & Communications
  • 412-268-1151

Carnegie Mellon University is playing a key role in Pennsylvania Governor Tom Wolf's planning efforts to re-open the state's economy.

There is no single dashboard tool that states can use to determine the best time to ease restrictions or re-open the economy, but the Commonwealth of Pennsylvania is partnering with Carnegie Mellon and other institutions of higher learning to make informed decisions based on the best data and science available.

This partnership, along with criteria developed by the Pennsylvania Department of Health, the Pennsylvania Emergency Management Agency and other commonwealth agencies and stakeholders in the areas of public health, economics, and emergency management, will help guide the state's economic and public health decisions moving forward.

In an April 22 media conference, Pennsylvania Gov. Tom Wolf commented on the collaboration with Carnegie Mellon on a dashboard to help determine when a region is ready to re-open and return to work.

To determine when a region is ready to re-open and return to work, the incidence rate of COVID-19 cases per capita will be evaluated and several public health requirements must be met. A regional assessment will measure COVID-19 cases per the population and will need to be an average of less than 50 cases per 100,000 individuals over the course of 14 days to return to work. The administration will work closely with county and local governments to enable the communities to re-open and transition back to work.

Additionally, to re-open a region, the commonwealth must ensure there is:

  • Enough testing available for individuals with symptoms and target populations, such as those at high risk, health care personnel and first responders.
  • Robust case investigation and contact tracing infrastructure is in place to facilitate early identification of cluster outbreaks and to issue proper isolation and quarantine orders.
  • Identification of an area's high-risk settings, including correctional institutions, personal care homes, skilled nursing facilities, and other congregate care settings, and assurances must be made that facilities have adequate safeguards in place, such as staff training, employee screening, visitor procedures and screening, and adequate supplies of personal protective equipment (PPE) to support continued operations. 

The commonwealth also will use a modeling dashboard under development and evaluation by Carnegie Mellon to take a regional and sector-based approach to re-openings, the easing of restrictions and response.

The administration will use this data-driven decision support tool to better understand the current health and economic status, as well as the inherent risks and benefits to re-opening certain businesses and industry areas. Using data that considers worker exposure and spread risks, health care capacity, economic impact and supply chain impact, the administration will prioritize re-openings where it has the potential for the most positive impact on the economy for workers and businesses, while mitigating risk to public health and safety.

"The purpose is to provide important information to the governor's team to make data informed decisions," said Ramayya Krishnan, dean of CMU's Heinz College of Information Systems and Public Policy and director of CMU's Block Center for Technology and Society. "For example, all indicators could point to opening a specific county, but other factors, such as population density around a hotspot, availability of supplies to ensure workers are protected, or Department of Health criteria could make the county unfit to open."

In order to arrive at results through this dashboard, the Commonwealth of Pennsylvania is providing access to data from several commonwealth agencies, including the departments of Labor & Industry, Human Services, Community and Economic Development, Revenue and Health.

The dashboard will help with questions such as: What will be the likely public health and economic implications associated with opening an industry? What impact might re-opening have on vulnerable workers and businesses? The data from the dashboard also will provide insights and recommendations at the industry and county level to inform state policy decisions.

The analysis will link together data sources to build an understanding of the current and real-time state of Pennsylvania's economy and the impact of the spread of COVID-19. The model will help to predict and understand what types of individuals, businesses, and industries will be more at risk, most vulnerable and impacted by COVID-19. The model will apply what-if scenarios that will allow the state to understand the impact of potential re-opening decisions, and to monitor changes over time and the impact of decisions.

"This has been an incredible opportunity and we appreciate the governor including us in this important work for the commonwealth," said Krishnan.

Members of the Carnegie Mellon team include:

  • Laurence Ales, associate professor of economics, Tepper School of Business;
  • Kasun Amarasinghe, postdoctoral research associate, Machine Learning Department, School of Computer Science;
  • Scott Andes, executive director, Block Center for Technology and Society;
  • Gary Franko, senior multimedia designer, Heinz College of Information Systems and Public Policy;
  • Rayid Ghani, Distinguished Career Professor, Machine Learning Department, and Heinz College; 
  • Jared Kohler, student, School of Public Policy and Management, Heinz College;
  • Ramayya Krishnan, dean of CMU's Heinz College of Information Systems and Public Policy and director of CMU's Block Center for Technology and Society
  • Tim McNulty, associate vice president of Government Relations;
  • Illah Nourbakhsh, professor of robotics, director of the Community Robotics, Education and Technology Empowerment (CREATE) Lab and associate director for robotics faculty at Carnegie Mellon;
  • Roni Rosenfeld, head of the Machine Learning Department, School of Computer Science, 
  • Randy Sargent, visualization director, CREATE Lab;
  • Richard A Stafford, Distinguished Service Professor, Heinz College;
  • Chris Telmer, associate professor of financial economics, Tepper School of Business;
  • Anne Wright, director of operations, BodyTrack;
  • Ariel Zetlin-Jones, assistant professor of economics, Tepper School of Business; and
  • Xuege Zhang, doctoral student, Tepper School of Business.

Other contributing members include:

  • Beibei Li, Anna Loomis McCandless Chair and associate professor of IT and management, Heinz College;
  • Lee Branstetter, professor of economics and public policy (joint appointment with the Social and Decision Sciences Department), Heinz College;
  • Jon Caulkins, Stever University Professor of Operations Research, Heinz College;
  • Karen Clay, professor of economics and public policy, Heinz College;
  • Baruch Fischhoff, Howard Heinz University Professor, Institute for Politics and Strategy, and Engineering & Public Policy;
  • Marty Gaynor, E J Barone University Professor of Economics, Heinz College;
  • Joel Greenhouse, professor of statistics, Department of Statistics;
  • Po-Shen Loh, associate professor, Mathematical Sciences, Mellon College of Science;
  • Dan Nagin, Teresa and H. John Heinz III University Professor of Public Policy and Statistics, Heinz College;
  • Rema Padman, Trustees Professor of Management Science and Healthcare Informatics, Heinz College;
  • Wes Pegden, associate professor, Mathematical Sciences, Mellon College of Science;
  • Lowell Taylor, University Professor of Economics, Heinz College;
  • Hai Wang, visiting assistant professor, Heinz College;
  • Peter Zhang, assistant professor of operations research, Heinz College