AI in Energy: What is stopping organizations from implementing AI?

We talk about AI a lot, about it’s potential to make our lives easier by revolutionizing the way we lead our lives and do business. However, what happens in reality? Are companies making use of artificial intelligence and its applications? If they are not, what are the roadblocks in front of enterprises that prevent them from making use of AI? According to a study by MIT, while 85% of executives believed that AI would help their companies gain competitive advantage, less than 20% had incorporated an AI system in some way, and just about 39% had a strategy for AI.

Why is a technology that promises to do so much being used so little?

AI needs to be trained on how to interpret and make decisions from the data it is presented with; therefore, the quality of data is paramount. This is true for the data used for training the AI and the data presented to the AI to make decisions from. Inaccurate data means the AI gets trained wrong or gives the wrong output. Introduction of an AI system will result in exponential growth in the data an organization handles; scale, security, reliability and uptime must be ensured through own storage or cloud.

Technology integration and scaling is another important piece of the puzzle. An AI system that is prototyped successfully in a lab still needs to be able to integrate well into the existing enterprise systems. There is often a huge divide between the digital capabilities supporting a pilot project and the capabilities available to fully scale and implement it. Implementation could be a long, complicated and expensive process or it might require a complete overhaul of the systems, processes, technologies in place.

Implementing AI requires skill-sets and capabilities that organizations might not already possess. Hence the need for finding the right talent and training existing talent to develop as well as properly use the new technology. Having AI systems also require their end-users to be involved from the beginning of the development. This means that to effectively develop an AI system that aims to improve customer experience, a team of customer experience experts who are familiar with all aspects of CX in the organization must be involved. Experts with specialized subject knowledge must be training and testing the AI system and its results.

We just investigated a few technological roadblocks; however, the real challenge is bigger than these.

Questions related to organizational strategy, policy, culture, technological capabilities, human resources, equipment and processes must be answered and clear road maps created, followed and reviewed at all levels and verticals of the organization.

This is a task that requires executives as well as technical experts to work in tandem from the bird’s-eye view down to the minute details.

Coming specifically to utilities, a survey by Roland Berger found that

83% of top European utility executives considered AI a high to medium priority for their business. However, 40% of them did not have an AI strategy nor defined targets.

While AI has applications at all stages in an organization’s operational and value chain, the way forward is not often as clear as executives want it to be, owing to the complexity of high-level strategic decisions need to be taken to choose the areas in which the organization should implement AI. Does your organization want to use AI to increase energy efficiency or to improve productivity or for predictive maintenance of equipment or to automate IT processes or to improve customer experience? The applications are endless; however, decisions must be made cautiously after taking into considering the priorities and the readiness of specific departments (in terms of skills, infrastructure) and budgets available. Only 2% of utilities in the Roland Berger survey thought that they had the people with the right skills to develop AI systems.

Utilities need to invest more in finding the right talent and training their existing personnel.

The challenges aside, results of the implementation of AI in energy have been more than promising. ABB’s AI-powered system is enabling industrial and commercial buildings to optimize their energy usage by reducing peak demand charges. This is accomplished by identifying patterns in the building’s energy consumption, while also factoring in weather forecasts and historical data. Siemens successfully deployed a neural network system that was able to reduce the nitrogen oxide levels from the combustion unit in gas turbines by 20 per cent. The results were visible within just two minutes after the AI system took over. General Electric’s Storm Readiness analytic combines weather forecasts, outage history, crew response, and GIS data to accurately forecast storm impact and prepare response crews and equipment ahead of time. These are just a few examples of the numerous applications of AI in the energy sector.

It sure is exciting times for artificial intelligence technologies, the applications could change the way how power and utility companies’ function. They would, however, need to change their risk-averse nature to embrace new technologies while strategically preparing their infrastructure, processes and people for the big changes AI can bring in. AI & ML : Data-driven Strategies for the Power & Utilities Industry brings together AI, data science experts and decision-makers from all verticals of leading power & utility companies in Europe to discuss strategic challenges and the way forward for making use of AI technologies to help their business become faster, smarter and more accurate and efficient. Connect with us to join the discussion.