Project Description:

While air pollution is the most visible environmental impact of transportation systems, water pollution and quality issues are also of great importance in the transportation and environment nexus. Specifically, transportation systems can affect water quality directly in many ways, including stormwater runoff, deicing chemicals, vehicle exhaust, oil spills, and other pollutants. However, the impact of transportation systems on water qualify is not well studied or fully understood. A significant challenge is the slow movement of ground water through aquifers and its long-lasting, detrimental effects on communities, aquatic life, and the overall health of the ecosystem. Addressing this challenge requires continuous and reliable data collection, as well as advanced data analytics techniques. The traditional method of manual data sampling and analysis is not sufficient. In this project, we will design and develop an AI powered remote sensing framework and associated algorithm for roadside water quality monitoring and prediction. We expect this effort will enable opportunities for causality analysis based on long-term historical data, leading to better understanding of the root causes of water pollution and potential strategies for preserving the water quality.

US DOT Priorities:

While air pollution from transportation is well-recognized, the project addresses the critical issue of water pollution caused by transportation systems, including stormwater runoff and deicing chemicals. This project directly addresses DOT’s research priority of “Preserving the Environment” by developing novel tools and technologies to measure, analyze, and mitigating water pollution caused by the transportation system. This approach, using historical data, aims to understand the causes of water pollution and develop effective preservation strategies through advanced research and technology development to combat transportation-induced water pollution.

Outputs:

In this project, we will design and develop a durable and cost-effective sensing system to measure a set of key parameters which can be used as indicators for water qualify, such as dissolved oxygen, conductivity, PH, and chemical pollutants. A prototype system design is shown in Figure 1(a), which involves configurable modules for power, sensing, computing, and wireless communications. We will perform an initial data collection in Salmon Creek watershed area, as shown in Figure 1(b). The interstate highways and state routes bisect the watershed, which makes it a perfect testbed for studying the short-term and long-term environmental impact of transportation systems. We will use the dataset collected along with open-source datasets to train machine learning models for pollution detection, characterization, and prediction. We will evaluate various algorithms and explore the potentials of performing causality analytics for mining the casual relations between complex parameters.

The outputs of this projects include both hardware and software systems, as well as an initial dataset for assessing water quality issues.

Outcomes/Impacts:

The proposed research will produce a remote sensing and data analytics framework for monitoring and predicting the water pollution caused by transportation systems. The project will produce new knowledge on the environmental impacts of transportation systems, as well as the potential causal relations. We expect these research discoveries will result in better understanding of the short-term and long-term environmental impacts of transportations systems, leading to a solid foundation for developing effective mitigation strategies. We will work with industry partners in both transportation and environment sectors and other stakeholders on potential technology transfer. In addition, the project will also provide valuable educational opportunities for graduate and undergraduate students, students from community colleges and K-12 schools, as well as working professionals.