Energy regulation plays a key factor for the electric mobility industry, especially when it comes down to trains. Next to that, congestion (blocking) of the energy net is a major problem in the Netherlands. Having insights in the energy usage of the network is crucial to help predict future demands based on events in the past.
We helped our client to get evidence-based insights to monitor and predict their energy usage of the power grid, which holds a maximum threshold. When this threshold is surpassed, a fine must be paid to the network operators (netwerkbeheerders).
Our challenge was to combine the scattered information over multiple systems within the organization, which led to a complex and mostly manual and labor-intensive process to validate the financial reports of the energy usage.
Therefore, we were asked to find a way to get more efficient insights into the current energy usage on the rail energy grid, and to estimate future usage, our client could save costs for having availability, while preventing the possibility of going over the maximum usage and paying the fine at the same time.
A toolbox was created that contains various tools for monitoring data ingestion and data processing. The development of a data platform is based on Azure components like Azure Data Factory, Azure Functions, and Azure Blob Storage. This platform is responsible for ingesting measuring data (energy consumption), invoice data, and business logic from multiple sources (five) and formats (more than ten). Additionally, a data processing pipeline was developed to aggregate and combine these diverse data sources into a functional database model. Finally, an introduced standardized development methodology based on test-driven development simplifies the implementation and transfer of the data platform and code to the internal development team.
The outcomes of our solution showed a positive impact on the business. Not only did we deliver a digital tool that combined all the different information channels, but it also lowered the manual labor for the coworkers and decreased the costs immediately. Some numbers at the end of this project:
- Removed manual labor it takes to process the invoices and controlling the data.
- Decrease of the overestimation with 10% on the required contract value with the network operator.
- Decrease of 5% of the fined over-contract cases.
- Decrease the time-to-insight from 10 days after the previous month to next-day insights.
- Less dependent on external consultants for the insights by automating away at least one FTE.