Background
For this project Dockside Data embarked on an ambitious project with a large, global automobile manufacturer. This manufacturer, operating multiple production facilities worldwide, faced a significant challenge. With the introduction of ISO 50001 regulation, they required a centralized solution for analyzing energy consumption across their diverse plants, while also determining the efficiency of their production processes. The complexity of integrating varied data structures from each location added to this daunting hurdle.
Challenge
The primary challenge was twofold:
- Data Warehouse Design: Designing and implementing a data warehouse (DWH) capable of ingesting and harmonizing IoT data from all manufacturing plants was crucial. Each location had its unique data format, and the sheer volume of data – several terabytes monthly – amplified the difficulty.
- Normalization for Efficiency Analysis: Beyond data collection, there was a need to assess whether the efficiency of the production had been reduced. This required the integration of normalization services using linear regression, creating and managing hundreds of custom models for each hierarchical level within the plants.
Solution
Dockside Data approached these challenges with a comprehensive strategy:
- Flexible Data Ingestion: The team developed a sophisticated data ingestion logic for the DWH, adaptable to various data formats from different plants. This was essential in creating a unified data environment.
- Advanced Analytical Logic with Normalization Services: Once data was consistently streaming into the DWH, Dockside Data introduced normalization services using linear regression. For each hierarchy level within the manufacturing plants, a custom linear regression model was created. This step was critical for analyzing whether the efficiency of production had been reduced, aligning and normalizing disparate data for accurate analysis.
- Virtual Meters for Complex Computations: Dockside Data engineered a system for virtual meters, capable of handling complex computations over vast datasets. This transformed the raw IoT data into actionable energy consumption insights, complemented by efficiency analysis.
Collaboration and Iteration
Continuous collaboration with the client using agile methods was a critical aspect of the project’s success. Dockside Data held several workshop meetings to deeply understand the manufacturer’s business requirements, which were pivotal in translating these needs into effective technical solutions.
Impact
The results were transformative, with the manufacturer now able to efficiently meet ISO 50001 regulation requirements, gain deep insights into energy consumption and production efficiency, and make smarter, data-driven decisions.
Conclusion
Dockside Data’s effective solutions enabled the car manufacturer to meet ISO 50001 standards and enhance energy and production efficiency. Their collaborative approach led to better data-driven decisions and significant operational improvements.
