Artificial intelligence holds promise in areas like predictive maintenance and tracking KPIs
Water and wastewater treatment processes have become more complex and sophisticated over the decades, but fluctuations in effluent quality and costs still happen, and their causes may be hard to pin down. Artificial intelligence (AI) can be a powerful tool in tackling many complex, nonlinear problems that arise in water and wastewater treatment.
Operational decision-making grows more complex and challenging along with the size of treatment systems and the number of variables that must be considered. As treatment systems grow to keep up with population size and urbanization, so do operational variables. At the same time, public and regulatory bodies have raised their expectations and compliance standards, but water utilities have only been able to consider data at a few points.
AI Applications
Recent research has focused on AI, machine learning (ML), and other optimization methods as a way to let programs develop themselves beyond their human programming. Among North American water utilities, 93% already incorporate supervisory control and data acquisition (SCADA), often with programmable logic controllers for automation. AI may run on top of SCADA systems, but SCADA simply executes preprogrammed commands. AI, on the other hand, actively learns, makes projections, optimizes, and offers new operational suggestions as it considers new data.
In wastewater treatment, AI models have become increasingly popular for monitoring water quality and optimizing process parameters, as well as for prediction and optimization during the treatment process. AI has also been adopted for a range of concerns, including:
- Evaluating the removal of conventional, typical, and emerging contaminants
- Optimizing process parameters and models
- Reducing membrane fouling
- Detecting and assessing pipe condition
Types of AI in the Water Sector
The three treatment AI model categories in use in the water sector are the artificial neural network (ANN), machine learning (ML), and search algorithm (SA). ANN roughly mimics the way the actual neurons in the brain work to understand complex nonlinear relationships. ML develops algorithms and predictive statistical models, and empowers computer systems to automatically learn from data without human intervention. Finally, SA searches for improvements using an accelerated version of the trial-and-error natural selection dynamic of biological evolution.
In advanced membrane systems, AI modeling can define optimal operating conditions for sensitive membranes to create effective fouling mitigation strategies, ensure higher effluent quality, optimize resource usage, and lower operating costs.
When applied to leak detection, AI can recognize data patterns that people cannot discern, producing better outcomes. Customer water use data reported by advanced metering infrastructure (AMI) has been analyzed by AI to create water conservation initiatives and generate water demand forecasts that are more accurate than linear ones.
AI has helped in the discovery of contaminant intrusion and other quality problems in water distribution networks. Problems in drinking water systems — such as drift from optimal performance — have been detected by AI, and it has successfully predicted bacterial hot spots and identified flaws in control strategies in distribution networks. Energy savings opportunities in pump scheduling have also been discovered by AI.
How Seven Seas Leverages AI
At Seven Seas, our engineers have been diligently developing and implementing AI-driven solutions, including smart maintenance systems and advanced data dashboards. These tools enable us to proactively monitor key performance indicators (KPIs), identify potential issues before they escalate, and optimize plant operations.
Seven Seas’ industry-leading record of 97% plant availability is one of the key focal points of our business, so our engineers are concentrating on using AI for predictive maintenance to further prevent downtime and outages. Also central to our business model is Water-as-a-Service® (WaaS®), which keeps our experts responsible for long-term plant operations and maintenance. By leveraging AI, we are able to analyze vast amounts of data from our plants and maintenance vehicles to anticipate equipment failures and schedule maintenance proactively. This approach also extends asset lifespan and ultimately lowers operational costs.
Our team of experts carefully evaluates potential use cases, assembles the necessary skillsets, and iteratively refines our AI models to achieve results. Through a combination of successful pilots and continuous learning, we are steadily advancing our AI capabilities to deliver tangible benefits to our customers.
New Avenues for AI in Water and Wastewater Treatment
Some researchers believe that AI will be applied to automated hydraulic model calibration, optimization of infrastructure planning, and emergency preventive repairs. Other research focuses on how AI’s data processing, pattern recognition, and optimization can aid in the creation of water-sector “digital twins,” which are digital, dynamic models of actual water systems that provide insights that can be acted upon in the real world.
Water-sector decision-makers, however, need more information on AI. Despite numerous benefits, AI presents many challenges, including the development of models, integration of knowledge, and ethical concerns that must be considered before real-world application, so utilities in the United States have generally been making the digital transformation slowly.
Seven Seas has already adopted some of these systems and is moving toward even greater AI adoption with maintenance-tracking programs that analyze data from plants and maintenance vehicles.
Accelerating AI Adoption in the Water Sector
Some projections say future AI expenditures for water and wastewater will approach $9 billion by 2030. While it is very easy to spend money on AI-related technologies, it takes precision and competence to focus on the most valuable business use cases to provide an actual benefit for the expenditure.
Seven Seas has learned a great deal from its AI pilots. Accelerating AI adoption requires the right team with the right skillsets to determine whether a project is viable or should be cut. If a project has promise, the right team can take the time to refine applications with iterative corrections until targets are hit.
Seven Seas predicts a continued increase in data dependency in the water sector that will require more sensors and data-gathering techniques, and we will continue to ride the AI wave. Contact our experts to learn more about AI options in water treatment.
Image Credit: peshkova/123RF
Matt Tesvich serves as the Global IT Director for Seven Seas Water Group. He is a seasoned manufacturing professional with over 20 years of experience in both high and low automation environments. He has been at the forefront of cutting-edge manufacturing technology, notably managing a $1 billion contact lenses plant outside of Atlanta, GA. As a former Global Manufacturing Digital Transformation Lead, Matt has driven innovation across multiple industries, including Medical Devices, Pharmaceuticals, and Utilities.