Utilities have a wealth of data at their disposal — everything from GIS and spreadsheet data to IoT sensor and SCADA data. Historically, a key challenge for utilities has been accessibility and usability of this data, but thankfully this can be overcome by some simple data and application road mapping, process improvement, and data integration tools. Once data accessibility is resolved, a new conundrum presents itself…what can we do with all that data?
There are of course straight forward answers such as visualization, reporting, and simple analyses (trending). However, with the world swimming in artificial intelligence (AI) headlines and companies adding “AI” features to nearly every application, how can utilities harness the power of these tools without getting bogged down with hype, niche solutions, and a low to no return-on-investment? This is where Data Science-as-a-Service (DSaaS) can help.
What is DSaaS?
DSaaS allows utilities to get data-driven insights leveraging advanced analytics models, without investing in fulltime, in-house data science experts, who can be hard to find, recruit, and maintain. Additionally, it does not lock the utility into a new, standalone application unless that is the best-chosen path.
It enables utilities to leverage their current data and systems, get tailored solutions from water domain specialists, and apply data science methods to their issues. Insights can be delivered in many forms – from tabular and geospatial deliverables to advisory systems integrated in existing software. Each utility can select what is best for them.
Benefits of DSaaS for utilities
There are multiple benefits of adopting DSaaS at a water or wastewater utility. The main benefit is connecting utility staff to water industry experts who also specialize in developing AI models to improve system operations.
Below are a few key areas where these models can be deployed to improve utility operations:
- Predictive maintenance (PM) for pumps, valves,etc.
- Workforce optimization aligning PMs for assets,crew training, schedules, and locations.
- Forecast operational trends such as influent flowand demand for treatment plants, network demand and capacity, and waterquality.
- Pump and lift station energy and conditionoptimization.
- Treatment process optimization includingchemical dosage, energy usage, process improvement, and peak demand/flow smoothing.
- Developing predictive models using utility data for predictive pipe breaks and overflows.
Given each utility’s systems, processes, and data quality are unique, adopting DSaaS allows you to step into AI pragmatically, at your pace, and focusing on your most critical needs.
Examples of DSaaS Applications
Predictive Maintenance
Reactive, proactive, and predictive maintenance are three distinct approaches for managing infrastructure. Reactive maintenance is performed after a failure has occurred, often leading to unplanned downtime and higher repair costs. Proactive maintenance, on the other hand, involves routine inspection and service to prevent failures before they happen, which can reduce downtime and extend the life of assets. Predictive maintenance takes this a step further by using data analysis and machine learning to predict potential issues before they lead to failure, allowing for more precise maintenance scheduling and potentially even greater cost savings and efficiency.
The application of each approach is typically dependent on the risk of the asset (using probability and/or consequence of failure). However, predictive maintenance is the least applied technique as data and analytic maturity can be a challenge. This is where DSaaS can help. Predictive models consume historical asset data along with many other sources to estimate future failures allowing better timed and scaled maintenance for prevention. For example, machine learning models can predict where and when dry weather sewer overflows may occur and help utilities optimize their cleaning and rehabilitation schedules both reducing events and saving budget.
Forecasting Operational Trends
Forecasting operational trends, such as network capacity during wet weather event and influent flow to a wastewater treatment plant, is another area where DSaaS can be useful for utilities.
By analyzing historical data, DSaaS can provide predictive models that can be used to predict process and capacity demands to help optimize operations. For example, in sewer networks, a machine learning model can predict the impact of wet weather on collection system performance to help the utility be proactive and minimize overflows. At a wastewater treatment plant, predictive models can be built to forecast influent flow and wastewater quality to the treatment plant thus allowing utilities to optimize processes and mitigate high flow disruptions in treatment processes. Furthermore, being able to forecast chemical and energy demands can help utilities plan better for the future and manage their limited resources effectively.
Empowering Communities with DSaaS
Data is at the core of many decisions, but the outcomes of these decisions are what truly impact communities and public health. This core to the Trinnex mission and is why our customers trust Trinnex with focused data science solutions to meet their unique needs. If you would like to get more insights into the topic, visit Trinnex’s DSaaS page.
