Research Interests

Modeling Human and Infrastructure Interactions in Sociotechnical Systems

The performance of urban infrastructure systems affects livability and well-being of communities, and the decisions to use resources and adapt uses affects the performance of infrastructure.  For example, electricity distribution, drinking water, wastewater, water reclamation, and stormwater systems can be affected by consumer decisions to adopt eco-innovations. These infrastructure systems, in turn, drive how consumers experience the delivery of drinking water and electricity and the removal of wastewater and stormwater. The goal of this research is to capture and simulate the complete feedback loop between infrastructure and human behavior. New models have been developed to simulate:

As an example of how this research was conducted, Dr. Berglund and her colleagues conducted a national survey to assess public perceptions about water reuse.  This data was used to encode an agent-based model to simulate opinion dynamics within a community, based on communication among households. A modeling framework that coupled agent-based modeling with EPANET was used to predict adoption as the initiation of new water reuse accounts and to assess the efficiency of water reclamation infrastructure expansion plans and the performance of pipe networks (Garcia-Cuerva2016; Kandiah2016Kandiah2017; Kandiah2019).

Developing Models to Understand and Mitigate Contaminants in Drinking Water

Water distribution systems are designed to deliver high-quality water to customers, but water quality is threatened by acute events, such as contamination, and through aging of pipe networks, which may allow metals to leach or result in contaminant intrusion through leaks. This research explores the reactions of consumers in acute events and simulated the effects of behaviors to reduce consumption on the total exposure within a population.  We developed new methodologies and computational frameworks to

For example, research conducted in the STSA lab developed new methods for managing a water supply contamination event and has investigated in depth how a community may respond dynamically to an event and change water demands based on exposure and warning messages. Results demonstrated that water consumers who react to water quality problems in their tap water can change the water flows in a pipe network to the extent that flow directions are reversed in some pipes, compared to the normal operating conditions.  These dynamics affect predictions about what segments of the population may be affected by the contaminant and what operations should be implemented to flush out a contaminant.

On-going research is exploring Bayesian Belief Network and machine learning models for predicting the presence of lead in drinking water systems (Fasaee2021).

Resilience and Regime Shifts in Sociotechnical Water Supply Systems

Complex interactions create challenges in managing water supply systems. Managers must adapt to changing climate, growing populations, and stresses imposed by other management units that share a common resource. This research explores how tipping points emerge in water supply systems.

  • Methods for calculating Fisher Information are implemented in a computational framework and applied to detect regime shifts in water supply data (Skarbek2019).

This research also explores the interplay among multiple managers and the emergence of water sustainability due to decentralized decision-making.

  • This research creates new insight about municipalities using water restrictions in a shared groundwater basin and the effects of decision-making on aquifer storage and water supply sustainability  (Al-Amin2018).

Smart Cities

Smart cities can use the Internet of Thing (IoT) technologies to operate more efficiently and conserve resources. In the water sector, the IoT connects personal smart devices and everyday objects, such as faucets, pipelines, and surface water bodies that are embedded with sensors, actuators, and network connectivity. Through the IoT system, user experiences and infrastructure conditions can be connected with automatic algorithms to take advantage of continuous data streams and send alerts to consumers or water managers. Utilities can take advantage of the IoT to identify leaks, encourage conservation, forecast water demands and quality, use crowd-sourced data to monitor water conditions, and warn consumers about water problems and hazards. This research is exploring a water smart city that operates its water infrastructure efficiently, provides a safe and reliable water supply, is well connected to its consumer base, and conserves water resources. Utilities in such a city can create novel programs by adapting their operations and planning practices in response to ubiquitous sensor networks and disparate data streams. Dr. Berglund taught a graduate level special topics course, Smart Cities, in Spring 2018, and on-going research is developing new methods and insight:

  • Machine learning models are developed to forecast hourly water demands at the account-level in real-time (Pesantez2020)
  • A new smart water grid is proposed to facilitate rainwater mico-trading among households (Ramsey2020)
  • Peer-to-peer markets for electricity trading are enabled by smart meters and blockchain (Monroe2020)
  • Community resilience can be improved through households that use social media to request help and respond to posts for help (DiCarlo2020; DiCarlo2021)
  • Our review of smart technology and infrastructure systems demonstrates the many opportunities for smart cities to improve delivery of resources and services across communities (Berglund2020)