Several research tasks were identified for the SWADE project: (i) determining interagency and community data exchange/sharing constraints and barriers; (ii) modeling integrated system structure and intersystem data interactions via the SWADE Ontology; and (iii) leveraging water systems data for planning, operation, and resilience, which includes operational workflow modeling and process mining, and supporting adaptive edge analytics through reinforcement learning. Initial work in each of these main research tasks were undertaken in different driving water system use cases.
Specific problems that are being addressed are listed as follows:
- Accurate Detection of Anomalies and Time-Varying Phenomena: Our focus problem revolves around dry-weather monitoring working with partners in Orange County, CA. Here, the goal is to identify the source locations and timing of effluents and contaminants into the stormwater system during dry, i.e. non-storm periods (and potentially illicit) flows. Current work in this direction is being done in the stormwater domain, and more details can be found here.
- Developing generalizable AI Models: Began to address the challenge of sharing solutions between different community agencies and stakeholders by developing a methodology for training and sharing data-driven monitoring analytical models. Our solution enables training these analytical models using data from multiple sources (agencies) in a distributed manner without the need for sharing the raw data which is always a big privacy concern for these agencies. Current work in this direction is being done in the wastewater domain, and more details can be found here.
- Initial SWADE Platform Architecture: A key aspect of the architectural model under consideration for SWADE is one that incorporates the separation of on-premise and in-cloud data for water agencies and organizations. The SWADE data exchange platform consists of two parts: the SWADE client that would be installed at the organization premises and the SWADE cloud where external data and analysis tasks are performed. Current work in this direction is being done in the drinking water domain, and more details can be found here.
Future and ongoing work will consider the following:
- Focused community interactions. In the coming year, working with Water UCI, an interdisciplinary center in the School of Social Ecology, we plan to continue the efforts in community engagement. In particular, we will conduct an additional set of focus group meetings and technology design plans in the context of the selected use cases to refine information needs and engage with our partners in platform and tool design. Our partners will also provide us with real-world datasets and testbeds that can be used for validation studies. Towards this end, we are developing one-on-one interactions with engineers and agency partners at the Los Angeles Dept. of Water and Power, Moulton Niguel Water District, Orange County Public Works (OCPW). CA Data Collaborative, CA Water Data Consortium, Irvine Ranch Water District.
- Extending tools/methods across water systems: The various research topics being studied (ontology design, data architecture for heterogeneous information representation, scalable data ingest, data cleaning and enrichment, process mining/modeling from agency logs, algorithms for combining model-driven and data-driven infrastructure resilience)are being developed and validated in the context of specific water systems. In the coming year, we will aim to study how the methods developed can be refined to apply across the multiple water systems – this will allow us to understand new cross-system exchange opportunities.
- Prototype SWADE platform design. We plan to enhance make significant progress in the SWADE platform and prototype design to merge community models, water infrastructure models and event models – this will require new techniques for incorporating community impact and large failures within the ongoing SWADE data architecture.
- Ontology Extensions. The development of the ontology started from the representation of the drinking/potable water system and the risk analysis task and will be expanded to the other water infrastructure (stormwater and wastewater) and types of analyses (e.g., water quality). We will enhance modeling of sensor datatypes (including water quality sensing parameters – turbidity, pH, DO, flow rates, pressure) for improved data management and exchange. In particular, we must extract metadata that is required for sensor (re)placement can be an initial step for health monitoring of the underlying system. In the coming years, we plan to expand our initial SWADE ontology to capture concepts across these different systems and focus problems, identify relationships and interdependencies, and conduct joint pilot studies with our community partners.