AI and the Energy Sector: Developing Tomorrow's Sustainable Technology

As artificial intelligence revolutionises more and more industries, its environmental impact is also increasing. Many view AI as an obvious solution for improved sustainability, but the truth is more complex and closer to a paradox where the climate solution itself risks becoming part of the problem. Large-scale AI models require significant computing power and energy, both during development and operation. For Gomero, the answer is therefore clear: sustainability must permeate the entire value chain, from development to implementation.

"When we develop our AI solutions, we consider sustainability at every step", explains Lennart Johannesson, senior developer at Gomero. "It's not just about optimising the electricity grid infrastructure for our customers, but also about how we build and maintain the software itself."

Through the SIPP platform, Gomero demonstrates how sustainability thinking can be integrated throughout the digital transformation. Three main aspects are in focus:

1. Energy-Efficient Coding

Gomero's development team actively works to minimise the resource requirements in their algorithms. Each line of code is optimised to reduce energy consumption in both calculations and data storage.

2. Long-Term Scalability

"We don't just build for today's needs", says Gomero's CTO Malin Giselsson, "but design systems that can grow and adapt without requiring a complete rebuild. This saves both resources and time."

3. Predictive Maintenance as a Foundation

By anticipating maintenance needs within energy infrastructure, both resource waste and the risk of failures can be significantly reduced. This is a central part of Gomero's sustainability strategy.

For the development team at Gomero, sustainable AI is more than just a buzzword. Lennart Johannesson describes the daily considerations: "We constantly balance performance and resource efficiency. Sometimes a minimal performance improvement can require disproportionate additional computing power. Then we must dare to choose the more sustainable path."

The team explores several innovative methods:

• Automatic downscaling of computing resources during low-load periods

• Intelligent caching to minimise redundant calculations

• Continuous monitoring of algorithm energy consumption

"A particular challenge is making predictive models both precise and energy-efficient", adds Lennart Johannesson. "It's about finding the optimal balance in data collection. We carefully analyse what amount of data is actually needed - we often see that fewer measurement points can provide the same quality of predictions. For example, we've received requests to collect three measurements per minute, but this would create a massive data surplus. In reality, three measurements per day are sufficient to achieve adequate precision. This saves energy not only in the data collection itself but also reduces the computing power required in our systems."

The energy sector faces an exciting future where AI will play an increasingly important role. Several trends are expected to shape development, including quantum computers which could revolutionise energy optimisation within a few years. At the same time, demands for transparency regarding the environmental impact of AI systems are increasing, and the EU's upcoming AI Act will likely include requirements for sustainability reporting for AI systems.

As the energy sector continues its transition, the need for sustainable digital solutions grows ever larger. Gomero's strategy demonstrates that it is possible to combine innovative technology with thorough sustainability thinking.

"We see this as the beginning of a new era in the energy sector", concludes Niklas Wicén, Product Manager at Gomero. "Where digital transformation and sustainability don't just coexist, but actually reinforce each other."