short to story
Innogy operates distribution grids, sells energy and generates most of its electricity from renewable energies. The company is primarily active as a grid operator and in energy sales. The company also plans, builds and operates electricity generation plants, primarily wind power plants (onshore and offshore), hydroelectric power plants and photovoltaic plants.
Goals & challenges
- Development of complex ML models
- Automated deployment of models via Amazon ECS
- Contribution of physical expert knowledge
- Coaching of internal data scientists
- AWS-Best-Practices
key points
- Use and integration of a variety of different data sources (geodata, sensor data, optical and infrared images, weather data, etc.)
- Re-evaluation of data collection procedures to correct identified problems in the data sets for the future
- Execution of data engineering tasks in the AWS
- Close coöperation with end users to ensure the usefulness and subsequent acceptance and use of the developed solutions from the outset
- Subtasks required a significant further development of the “usual” algorithms (e.g. feature detection on low-structure surfaces)
services accomplished by synvert saracus
- Development of various use cases (hydropower plants, wind turbines and e‑mobility)
- CRISP-Workflows
- „The Data Science Process“ – Methodology (Business Understanding, Data Understanding, Data Preparation, Modellbildung, Feature Engineering, Evaluation, Deployment)