News & Announcements

Ohio Department of Transportation Funding Will Improve Emission Monitoring Around Construction Sites

Mehdi Rahmati, Emmanuel Kidando, and Josiah Owusu-Danquah will utilize swarms of drones and machine learning to provide real time, precise environmental monitoring

April 2024 Research NewsletterDr. Mehdi Rahmati, an assistant professor in the Department of Electrical and Computer Engineering (ECE), alongside co-PIs Dr. Emmanuel Kidando and Dr. Josiah Owusu-Danquah, both assistant professors in the Department of Civil and Environmental Engineering (CEE) have been awarded a one-year, $100,000 grant by the Ohio Department of Transportation to combine advanced unmanned aerial vehicle (UAV or drone) technology, cutting-edge machine learning, and intuitive visualization methods to improve monitoring of greenhouse gas (GHG) emissions around construction sites.

GHGs are a global concern due to their significant contribution to climate change.  Road construction and maintenance are notable contributors to GHG emissions, so the research team is developing a novel method to monitor key pollutants Nitrogen Oxides (NOx), Carbon Monoxide (CO), and Particulate Matter (PM10 and PM2.5). Utilizing a Deep Neural Network (DNN) based Reinforcement Learning (RL) method, a swarm of UAVs equipped with sensors will be flown and detect pollutants around the construction site, preemptively adjusting their positioning and operational parameters based on environmental factors. RL specifically aids in recognizing real-time changes in gas concentrations, adjusting the UAVs' routes and altitudes for optimal sensing. For instance, if a particular area shows a sudden spike in emissions, RL can direct UAVs to focus on that region, ensuring high-resolution data collection. Further, by understanding patterns and learning from past data, the UAVs can anticipate emission hotspots and allocate resources accordingly.

The combination of DNN and RL ensures that UAVs don't just collect data but do so intelligently, maximizing precision and minimizing data gaps, leading to more accurate and comprehensive environmental gas monitoring.