Anomaly Detection on Energy Consumption Data

Throughout the semesters, we have explored several approaches, including generative models and statistical models. Ultimately, we selected a statistical model due to its outstanding performance and ease of interpretation.

The final delivery included a complete end-to-end pipeline, from data collection and preprocessing to model application. Results are presented via an interactive dashboard capable of auto-generating reports. Additionally, we provided a highly configurable command-line interface (CLI), enabling practitioners to efficiently experiment with and run different model configurations.

Our solution significantly reduces manual effort, empowering practitioners to quickly detect anomalies, better interpret energy usage trends, and conduct more thorough, in-depth analyses.

  • Kjeldsberg Eiendomsforvaltning AS is a property management company focused on creating safety for owners and comfort for tenants. With its headquarters in Trondheim, Norway, the company offers a range of services aimed at optimizing energy use, environmental sustainability, and overall property management.

    The company provides tailored property management solutions to ensure all aspects of a property are maintained with high standards of quality and security. They have extensive experience in fire safety and security, and their technical team is well-versed in managing all phases of building projects, from planning to execution. Additionally, Kjeldsberg offers financial and administrative management services, including budgeting and accounting.