
The COVID-19 pandemic was exacerbated especially in the early stages by a lack of real-time and predictive data collected on the extent, location and spread of the disease. One Durham-based startup saw this information gap as a call to action.
Through a grant received from the National Science Foundation last year, Durham-based research and development startup Geometric Data Analytics (GDA) is developing a platform to model disease spread and provide real-time identification of infectious disease outbreaks, focusing initially on COVID-19.
GDA—which is currently participating in the eighth cohort of the RIoT Accelerator Program (RAP)— was founded in 2012 by John Harer, Professor of Mathematics and Electrical and Computer Engineering at Duke.
GDA grew out of research performed by Harer, who is GDA’s CEO and a pioneer in the field of topological data mathematics. The 10-person team applies cutting-edge mathematical techniques with other approaches in the fields of machine learning, data fusion, shape analytics and edge analytics to solve complex data-analysis problems.
The research startup does mostly government contract work and has been contracted by 10 government agencies so far including the Department of Energy (DOE), Department of Homeland Security (DHS) and the Defense Advanced Research Projects Agency (DARPA). GDA also works with non-governmental companies in various industries such as public health, agriculture, supply chain and logistics, and ecological monitoring.
In April, GDA was awarded an SBIR grant from the DOE to continue their research on another front: analyzing the roots of trees, vegetables and plants with plant biologist Dr. Anjali Iyer-Pascuzzi at Purdue University. The root analysis is being paid for out of the $110 million that the DOE awarded to 102 small businesses across 24 states that are pursuing research on sustainability, clean energy and climate solutions.
Root systems play a critical role in ecosystems and agriculture, but they are hidden and difficult to study. The image analysis technology that GDA is developing automatically processes pictures taken of root systems underground to make it faster and cheaper to measure roots and their growth.
The Root Shape project and COVID-19 disease monitoring platform are the two primary projects GDA is currently working on. For the latter, GDA is applying data fusion approaches to leverage data from multiple sources relating to infectious disease incidence, transmission risk, and sub-population interaction to generate predictive models on disease spread.
Ashlee Valente, senior scientist at GDA, is leading the public health monitoring project. She said GDA was part of a wider response by other agencies and institutions at the beginning of the pandemic to solve the lack of real-time data being used for tracking and predicting disease outbreaks.
“When the pandemic hit, there were all of these calls for health solutions, data solutions and anything people could do to combat the situation, and we wanted to contribute in any way we could,” Valente said. “So when we saw the NSF call for proposals specifically around addressing the lack of data, our idea was to apply our data fusion capabilities to solve that problem. Our hope is that we can use our data fusion capabilities to make a difference in the public health monitoring and response space, and already we have a more robust public health monitoring infrastructure than we did a year ago.”
Valente is also leading GDA’s participation in the RIoT program. She said their decision to apply to RIoT was motivated by their effort to expand the scope of their work.
“The pandemic has been a catalyst for us to explore other industries and areas where we can apply some of our data-fusion capabilities to solve real-world data challenges,” Valente said. “And being in the RIoT accelerator program is already pretty helpful for giving us an opportunity to really explore some of those new ideas.”