RSISSC REU Site Project Themes

Assessment of hazard risks to aging infrastructure systems that analyzes metrics for resilience and sustainability over extended decision horizons. These efforts will support infrastructure asset management decisions, e.g., project prioritization [3,4].

• Participants will learn risk-based performance assessment over the life cycle of systems and applications for managing wastewater or bridge networks with emphasis on analyzing the distribution of risks in disadvantaged urban communities. Participants will use MATLAB for modeling and data analysis.

Enhanced integration of industrial by-product materials as mix components for concrete and asphalt to provide low-cost and low-carbon paving materials for replacement and rehabilitation of aged and degraded highway pavement near cities around the nation [5].

• Participants will create concrete specimens with recycled materials and test their macro-scale mechanical properties and durability. Deployments in in-service conditions will be used for field performance evaluation.

Detection of bridge motion patterns with non-invasive visual monitoring of bridge motion using digital imaging and analysis.

• Participants will install a camera-based survey system on campus bridges with GPS-benchmarked locations and use photogrammetric approaches to convert recorded images at the pixel level to absolute coordinate locations. Data analysis techniques are used to detect anomalies in the motion patterns.

Identifying trends in the movement of goods, vehicles and people in indoor spaces by using geotags (location with a time stamp) to contribute to decision-making in smart city management.

• Participants will acquire rich sensor data streams [6] and then use machine and deep learning to reconstruct the object space in terms of geometry. Participants will perform semantic labeling of objects and gain understanding of new technologies for mapping civil infrastructure systems [7]. 

Improving autonomous vehicle navigation with low-cost imaging and 3D reconstruction to provide semantic models (location model of objects labelled by role/function) in near real-time.

• Participants will apply new concepts of multi-view geometry in photogrammetric mapping with images from vehicle-mounted ‘GoPro’ cameras and high-resolution commercial satellite images [8]. Semantic segmentation with deep-learning methods will be performed to enhance geometric models of vehicle surroundings by providing object identification.

Centering transit planning around traveler-centric metrics to guide returns on investment in efficient rapid transit systems by characterizing the quality of people’s overall (“door-to-door”) travel experience [9].

• Participants will use a dynamic traffic assignment framework, FastTrips [10], to simulate a variety of service levels on a major transit network, to assign passenger streams accordingly, and to extract passenger-focused planning metrics.
• Participants will work with large-scale data sets and data analytics approaches to make direct connections between big-picture policy goals and day-to-day operational outcomes.

Application of ultrasound technology to reduce the growth of harmful algae so possible toxin production in reservoirs does not impact drinking water treatment plants.

• Participants will conduct fieldwork to assess the effectiveness of buoy-mounted ultrasound transducers. Complementary laboratory observations of cell physiology will be used to better understand susceptibility of cells to chemical attack through byproducts of ultrasound cavitation in water.

Optimization of low-energy ultraviolet light emitting diodes for destroying pathogens in water systems has many applications for point-of-use treatment or small water systems and in-pipe applications.

• Participants will perform fundamental rate kinetics experiments, coupled with genomics techniques to identify cell susceptibility to damage. These observations will be used to develop of pathogen species-specific models to improve efficacy of ultraviolet light dosing.

Characterization of humidity conditions that trigger excess growth of fungi (mold) in house dust [11,12].

• Participants will use bioinformatics to examine existing DNA sequencing data [13] to characterize the diversity of fungal species known to contain human allergens. They will next extract house dust samples to sequence the DNA and characterize fungal species present to better understand the microorganisms that exist in homes and how they may impact our health.

Deployment of low-cost particle sensors to provide high-resolution and near real-time data for outdoor air quality [14].

• Participants will couple the collection of primary air quality data using low-cost sensors [15] with publically-available data from networks designed with consideration of health outcomes, environmental justice and other factors.
• Participants will engage with planners, statisticians, environmental health scientists, and regulators.

Implementation of data-driven control of indoor ventilation to balance occupant comfort and health, building energy use and grid demand. Participants will use state-of-the-art building simulation software, such as CONTAM and EnergyPlus, to analyze data-driven control strategies for building ventilation systems.

• Participants will be able observe direct feedbacks to indoor air environments that are critical to the health, comfort and productivity of inhabitants.

Enhancing the fidelity of large-scale physics-based models of overhead utility infrastructure to improve predictive capabilities for the magnitude and locations of power grid outages subject to extreme natural hazard events.

• Participants will employ data mining strategies with public data sources along with existing physics-based fragility models to examine the relationships between performance and physical features of the power infrastructure system and the environment and to draw recommendations for strategies to improve the resilience of the power grid.
• Participants will contribute to strategies for bridging gaps between expensive physics-based and unspecific statistical model representations for assessing infrastructure reliability.

Buffering the time-shifting generation of solar and wind energy generation through the novel use of carbon dioxide as a geothermal heat extraction fluid.

• Participants will use simulation models that couple energy load curves with renewable energy generation patterns to investigate the interplay of elements in this coupled system with the goal to avoid carbon dioxide emissions.
• Participants will examine the implications of these simulations on energy management policy.