The Power of Data for Smarter Utilities Asset Management

The utilities industry faces unprecedented challenges. Aging infrastructure requires immense investments in updating all components of transmission and distribution lines. At the same time, there’s a movement towards clean energy – which requires enabling electrification and reducing emissions.

Both are key areas of focus for many utilities organizations as they continue to build the grid of the future with a focus on delivering clean energy while increasing energy production. Optimizing asset management continues to be a priority for many utilities.

With Maximo as the hub for many enterprise asset management (EAM) solutions, utilities typically hold an immense amount of asset data. However, using this data to make daily decisions around managing assets can be challenging. Depending on where utilities may be in capturing, using, and analyzing their data there is an opportunity to unleash the power of data – using process automation tools and advanced data science technologies.

Leveraging Data for AI-powered Insights

With a suite of data visualization tools, predictive models, optimization engines, and text synthesis and generation capabilities, modern AI is helping to drive utilities organizations toward more efficient, data-driven decision making.

While there are endless opportunities for AI to generate a significant return on investment, there are a few key areas to drive initial results.

Automated Work Scheduling

Often times, maintenance teams face challenges with work scheduling. Specifically, they rely on time-consuming, manual process to create schedules that assign workers to work orders on a daily basis. The process is very dependent on the experience and preferences of the schedulers and further complicated by the many unwritten rules of scheduling work – including work order priorities and requirements, worker skills and availability, and scheduled outages; to name a few.

By leveraging automated work scheduling, it can simplify the process while increasing scheduling quality. This helps to drive efficiency across the team to get the work done.

Instead of reviewing work order information in multiple reports, and then comparing the work with available resources, the scheduler could initiate the process via an automated work scheduling platform. In doing so, the work management software pulls in work order and resource data. It can then share this information with an AI-driven optimization engine to generate and propose the best possible work schedule; which users can then view, accept, and adjust within the work management software.

Shifting to an automated work scheduling process offers several benefits, including:

• Eliminating manual tasks so employees can focus on more meaningful task
• Improving schedule accuracy by over 25%
• Propelling maintenance organizations to proactive, rather than reactive

Preventive Maintenance Frequency Optimization

Finding the correct scope and frequency of preventive maintenance (PM) tasks is challenging for all organizations. After all, each piece of equipment has different needs and requirements.

Applying a one-fit-size-all maintenance schedule often leads to equipment being maintained too much or too little. Both create challenges – too much preventive maintenance can be a waste of resources and create additional maintenance issues; too little can lead to high corrective maintenance costs and reliability issues.

With AI, utilities can optimize PM frequency by leveraging the asset and maintenance history data. This information then is translated to machine learning models to predict labor costs and hours. From there, a detailed recommendation of a PM frequency to help minimize labor hours and reduce excess costs due to over-maintenance.

Capital Portfolio Planning

Given the aging infrastructure for utilities in the United States, combined with utilities’ quests for a greener future, capital equipment requires substantial investments.

Often times, capital portfolios are planned based on subjective parameters and requirements; for example, projects may be approved because of internal champions for those efforts. This can raise challenges when evaluating the effectiveness of a capital portfolio plan and ensuring that it meets the values and needs of the organization.

Organizations manage long-term asset management processes by leveraging data to rank proposed projects and make data-driven decisions based on organization-defined parameters. This technology, along with AI, can optimize where capital budget spend should occur.

First, the system prioritizes projects, based on the value of projects and the risk of deferral. This informs how to optimize the plan to maximize that value and minimize deferred risk, which results in a comprehensive capital portfolio calendar.

Through this approach, the capital planning team gets the ability to define objectives, rules, and constraints to maximize investments today and in the future. This results in improved capabilities to plan faster and smarter, accounting for all practical constraints and considering all projects consistently and objectively.

Exploring “what-if” scenarios and adjusting without starting over becomes an easier task, giving end users the ability to see where they can maximize investments and results.

And, as organizational priorities continue to drive capital portfolio planning forward, using a solution like Endevor can better align project work with company priorities – while providing the documentation necessary to report on what’s happening and why.

Dissolved Gas Analysis Alarm Management

Online Dissolved Gas Analysis (DGA) can generate hundreds of alarms per day, which can be challenging to quickly review and determine the appropriate course of action.

DGA monitoring benefits greatly from AI. A tuned machine learning model can help identify the alarms that require attention – classifying them based on the dissolved gas behavior and providing a recommended action.

To do so, gas-in-oil time-series samples must be captured and leveraged in its data science technologies to classify each sample and outline recommended actions. Machine learning classification algorithms help train a best-in-class machine learning classification model.

The AI-powered approach enables teams to analyze more DGA samples, while making it easier to identify important alarms and flag for resolution. As a result, the team can catch more issues early. A seamless integration with asset management software routes alarms to those responsible for fixing the issue, and creates work orders and other documentation as necessary (some of which can be further automated via Endevor’s platform).  This integration gives a complete view of DGA results alongside other insights in the platform.

Beyond these top use cases, the presence of a broad, high-quality data set enables countless other opportunities that can benefit the organization.

The true advantage of utilizing a process automation platform powered by AI lies in its limitless potential to optimize, automate, and enhance reporting.

Interested in learning more about the power of AI-powered insights for utilities organizations? Tune into our on-demand webinar.