(156b) Keynote Talk-Machine Learning Based Software Tool for Efficient Wastewater Asset Management

As per the statistics in 2017¹, more than 27 million Americans are exposed to water systems in violation of the health standards established in the Safe Drinking Water Act (SDWA).² The American Society of Civil Engineers (ASCE) infrastructure rating systems have rated the Unites States’ Water (W) and Wastewater (WW) category as D and D+ respectively³, which is alarming and directly correlates to the issues of Water Quality, Equity, and Public Health. Thus, there is an immediate need for convergent thinking and holistic solutions for enhanced W&WW management systems that concurrently address infrastructure improvements, as well as enable significant revisions in the water and wastewater quality monitoring, treatment, and management systems, to meet the SDWA guidelines.

Most water utility companies use a reactionary or subjective approach⁴ for safeguarding water quality and infrastructure. The problem with a reactionary (wait-watch-act) approach is that problems usually cannot be solved until a failure or hazardous event occurs. Once an issue arises, significant amounts of manpower and capital are needed to access and repair the infrastructure failures⁵, and, at the same time, it is important to minimize any negative impacts on the surrounding environment and the quality of water in the distribution systems. Thus, the reactionary approach leads to large capital losses and negative environmental impacts due to the lack of a backup system or plan of action. The new enabling ML-based software tool developed by our team will serve municipalities, utility companies, and facilities for informed decision-making and timely management of their wastewater infrastructure. We have combined methods from data collection, analysis, mathematical modeling, artificial intelligence, machine learning⁶, and optimization⁷ to develop the software tool which will perform vulnerability and risk assessment of existing utilities and infrastructure and create plans of action towards resilience and manage the wastewater collection systems effectively.

References:

(1) PolicyLink. Water, Health, and Equity - Clean Water for All; 2017; pp 1–32.

(2) US EPA, O. Safe Drinking Water Act (SDWA). US EPA. https://www.epa.gov/sdwa (accessed 2021-02-08).

(3) America’s Infrastructure Grade. ASCE’s 2017 Infrastructure Report Card.

(4) Barles, R. Drinking Water Infrastructure Needs Survey and Assessment. 76.

(5) Aging water infrastructure ‘nearing the end of its useful life’ - Investigate MidwestInvestigate Midwest. https://investigatemidwest.org/2014/06/23/aging-water-infrastructure-nea... (accessed 2021-02-09).

(6) Kumar, A.; Flores-Cerrillo, J. Machine Learning in Python for Process Systems Engineering. 352.

(7) Roshani E.; Filion Y. R. Event-Based Approach to Optimize the Timing of Water Main Rehabilitation with Asset Management Strategies. Journal of Water Resources Planning and Management 2014, 140, 04014004. https://doi.org/10.1061/(ASCE)WR.1943-5452.0000392.