Does big data hold the clue to traffic fatalities?
Connecting state and local government leaders
Researchers from Penn State are developing a high performance computing-based framework to evaluate accident traffic data from police reports across California to find patterns related to deaths and severe injuries.
Fortunately, severe injury and death are relatively rare results of today's traffic accidents. Because those few incidents make up a small sample in the total number of accidents in any given locality, it’s difficult to draw any data-driven conclusions about the causes of such accidents. But if there were a way to analyze a substantially larger number of accident reports, clues to the causes of traffic fatalities might be revealed.
Two researchers at Penn State University are developing a high-performance computing-based framework to evaluate accident data from police reports across California to find patterns related to deaths and severe injuries.
Kamesh Madduri, assistant professor of computer science and engineering, and Venky Shankar, professor of civil engineering, aim to create software that can read the large datasets of traffic information and possibly uncover the common factors of deadly accidents.
Shankar called the research “fundamental in terms of the computational methods being put to use to extract patterns relating to fatalities and severe injuries. It is a tool for discovery and has the potential to provide unique insight into the contexts in which fatalities and severe injuries occur.”
“Statistical analysis of large accident datasets involves hundreds of variables, and the state-of-the-art in computational methods for estimating complex fatality models from large datasets is nonexistent,” Madduri said. "We’ll develop software to do this."
However, the idea is preliminary and high-risk -- no one has explored this area of data, Madduri said. “We'll be working with large, multivariate datasets and there is a substantial gap between theory and practice in this area.”
Once the software is developed that can indicate the factors associated with fatal and severe accidents, the researchers plan to apply it to other states’ traffic data. “If we can create a tool that is fast, and provides quick-response analysis for national safety policy, we can move one step closer to reducing fatalities and severe injuries by one half,” Shankar said.