FROM ASSUMPTIONS TO FACTS, SHIFTING THE CONVERSATION TO BETTER MANAGE TIME, EFFORT, AND MONEY

A Croutear-Foy, AquaWatch Solutions

ABSTRACT

In the realm of stormwater management, integrating Visual AI with the analysis of physical water quality parameters presents an approach that is more and more prevalent, but brings with it as many challenges as it solves. This presentation will explore the benefits and applications of this innovative methodology, focusing on how visual AI technologies coupled with water quality monitoring can revolutionize stormwater management practices and target intelligent, fact based investment strategies.

The crux of our approach lies in the utilization of Visual AI to provide real-time, accurate observations of stormwater systems and using machine learning and large language models to integrate the data provided by continuous physical water quality information. This technology enables us to visually monitor critical aspects such as trash rack management, underground gross pollutant traps, and combined sewer overflows. By correlating these insights, we gain a comprehensive understanding of the stormwater system's health and functionality.

Case studies will be presented to illustrate the effectiveness of this method. These include managing trash racks more efficiently by identifying clogging events in real-time, monitoring underground gross pollutant traps to assess their effectiveness, and observing combined sewer overflows to mitigate pollution risks. Each case study highlights how visual AI enhances our ability to respond promptly and accurately to various challenges in stormwater systems.

We will also look at how we can leverage the data that we already have access to, delve into how large-scale data, collected through visual AI and water quality sensors, is used to model and predict system behaviours based on real time understanding, rather than intermittenet and incomplete grab samples. This data-driven approach married with practical analysis allows for the development of management strategies that are not only based on factual evidence but also tailored to specific environmental conditions and challenges. The predictive models that can be generated from this extensive data aid in optimizing investment decisions, ensuring that resources are directed towards the most impactful solutions.

It's not enough to work in silos, we need to collaborate and be brave in sharing information. Providing detailed, real-time insights into the physical state of water systems, their infrastructure, and their quality, this method allows for more informed, effective, and sustainable management practices.