Cognitive AI is making utilities smarter

Dec. 29, 2020
As the brainpower retires, artificial intelligence is emerging as a critical enabler.

By AJ Abdallat, CEO and founder of Beyond Limits

The power industry is at a crossroads.

The demand for more efficient, large-scale power generation is higher than ever, as the world's cities are becoming increasingly populated and more connected to the grid. Historically, energy and utilities companies have hired the most experienced facility operators to manage complex facility processes successfully. These operators are seen as the best for complicated jobs because they can draw upon their expertise to achieve high operational efficiency and reduce risk to personnel and equipment.

However, the talent pool for experienced operators is shrinking with Global Energy Talent Index (GETI) reporting that 48% of power professionals are concerned about an impending talent emergency. In addition to the talent crisis, the energy industry is also looking at new solutions to address the global demand for safer, cleaner energy plants and distribution networks. As a result, utilities and energy providers have begun to turn to artificial-intelligence solutions to increase operational efficiency while reducing environmental impact and preserving the knowledge and expertise of an imminently retiring workforce.

Saving veteran knowledge

The number of experienced operators is waning. Their years of experience and essential knowledge cannot be lost upon retirement. These senior-level individuals have decades worth of knowledge too valuable to be squandered and forgotten to the annals of history. The good news is that energy providers can now use cognitive AI to help record veteran operator knowledge and distribute that knowledge, best practices, and business logic to less-experienced operators.

Today, modern developments in technology and digitalization software, such as cognitive AI, have the ability to collect, codify, and distribute that essential knowledge and expertise so it can exist in a perpetual state across an organization or industry. Advanced systems of this nature can also combine archives from past operator decisions with near real-time facility data gathered from multiple sensors to provide recommendations on next steps, including the context for why and how the system came to a decision. Cognitive AI captures data from those sensors, which track system levels such as vibration, acceleration, acoustics, temperature, pressure, etc., while leveraging human experience to deliver actionable information.

These transparently auditable, AI-based recommendations result in faster decision-making and more optimal actions, boosting operational insights, improving operating conditions, enhancing performance at every level, and ultimately increasing profits as a result. Such capabilities also make it possible for newer, less-experienced platform operators to be introduced to historical knowledge sooner. This way a new operator may acclimate more quickly, thus accelerating time-to-training while mitigating risk.

Efficient operations

Cognitive AI is being used to provide each energy facility’s and utility plant's workforce with a more holistic view, enabling personnel to see both the big picture and any minuscule detail that, if altered, may inadvertently affect other towers or operating systems within the plant. The AI system is able to continually evaluate the current state of the plant, recent production, performance history, and load demand to provide decision-makers with recommendations, including:

• Identifying and predicting demand and usage patterns

• Bringing turbines, generators, and other assets online/offline

• Adjusting asset operations to account for variable ambient conditions and other constraints

• Increasing asset performance and meeting demand by following the load while running assets in the most effective way possible

• Identifying tradeoffs between fuel cost, asset health, and responsiveness to the grid

Not only do these recommendations raise the day-to-day level of efficiency of utilities, but they have also been shown to lower unexpected and costly maintenance expenditures by virtue of the system constantly analyzing the non-stop stream of data. 

Streamlined maintenance

By monitoring energy and utility systems’ states of operation, cognitive-AI solutions are also able to identify performance trends and predict risk increases for asset degradation or failure. These systems can issue recommendations on appropriate mitigation and maintenance plans. Within this context, such AI has the ability to:

  • Identify baseline behavior and anomalies through the ingestion of sensor data (e.g. vibration and temperature)
  • Use historical data from sensors, events, and environmental conditions to identify patterns and trends that can necessitate more frequent maintenance
  • Collect and analyze images and data to identify wear and damage such as rust, coking, hot spots, warps, degradation, fractures, and more
  • Make recommendations on best steps upon identification of likely defect and damage

These capabilities are particularly critical now, as utilities providers are taking an increasingly conservative approach to hazardous situations, especially since PG&E's landmark criminal conviction in March for its faulty power line igniting California's deadly 2018 Camp Fire.

The present and future of AI

The capabilities of AI in the utilities space have only scratched the surface of the technology’s potential impact. So much more can be possible if/when utilities companies are willing to make investments in their technological advancement; these investments will help amplify decades of existing industry-operator intelligence and intuition.

The benefits of cognitive AI are already being proven in many areas both within and outside of the utilities industry. AI will also give consumers more power and influence over the amount of energy they budget for and utilize. Energy users will even be able to contribute to the power grid themselves by selling power back to their energy provider and negotiating prices.

Through AI, energy providers are starting to tap into another level of efficient energy production, and on top of that, they are contributing to the larger global goal of producing cleaner, less environmentally impactful energy.