New technologies and IT solutions make a significant contribution towards ensuring RWE is prepared for the challenges of the future. To be able to generate the desired added value, information in the form of analyses and modelling is now more visible than ever before. Digitalisation makes processes transparent, in power stations every bit as much as on the trading floor or in projects with partner entities.
RWE is proud of the fact it has extensively integrated several hundred apps and various cloud services into its existing business processes. This opens up many different opportunities for company employees.
Algo Trading makes it possible to respond in the most effective way to movements on the energy market at any time. Although this is a relatively new concept for many market players, RWE Supply & Trading has been making use of this process in its day-to-day business since 2015. The digitalised and automated approach to trading enables the trading company to achieve the best results on the market – continuously, efficiently and reliably. On the intra-day market in particular, where prices change in seconds, the complex algorithms developed in-house achieve crucial advantages. In addition to the flexibilities of RWE’s power stations, many customer positions are also managed using these algorithms. That means customers benefit from a constant proximity to the market and efficient trading decisions, while keeping their own manual input to a minimum. To get the most out of the potential offered by automated trading, the experienced traders at RWE Supply & Trading work every day to improve the performance of the algorithms and adapt them to rapidly changing market conditions.
The innovative web-based service “Easy Commodity Trader” (ECT) offers B2B customers of RWE Supply & Trading a professional solution for bilateral commodity trading. This portal, developed in-house, gives customers easy, fast and reliable access to the key European electricity and gas markets. Thanks to the customisable trading screen, users can easily trade in both standard and non-standard products from their screen to suit their own preferences. ECT is easy to integrate into existing systems via an API interface, and is free to use. Two-factor authentication and the use of a session token guarantee that trading remains secure, while providing support for a professional trading experience.
“Lead in Data” (LiD) is a cloud-based data platform that allows an entirely new form of data management at RWE Supply & Trading. This future-proof, scalable platform is perfectly adapted to the company’s commercial strategy and is able to provide efficient access to data and facilitate working with extremely large data models. To make the best use of the full potential offered by the platform, both a special training course and a user community have been developed for the more than 350 individuals who use it every day. The ability to work collaboratively and access to a central data catalogue make LiD the ideal solution for multi-team communication within the analyst community.
Machine Learning is revolutionising document processing at RWE Renewables Europe & Australia. The “Digital Transformation & Innovation” team (IT/OT) has set itself the goal of optimising and automating document processing in order to reduce the time otherwise required for manual and monotonous processes.
Machine Learning has its origins in Artificial Intelligence, and enables human abilities such as logical thinking, learning, planning and creativity to be imitated in various fields of application and then further expanded using data algorithms.
Machine Learning is used within the RWE Renewables Europe & Australia IT/OT team in various areas of the company in order to automate and optimise document processing.
Working with third-party companies creates system and process breaks in some departments that make it difficult to share data automatically. Data is often transmitted in the form of PDF documents or even in paper form, and so processing that information requires a high level of manual effort. In addition, the task of processing these documents manually is often monotonous and time consuming. Using Machine Learning makes it possible to speed up and simplify this process, and also minimise susceptibility to errors.
Initial tests have been performed at RWE Renewables Europe & Australia to train Machine Learning models in text and table structures from documents such as letters or reports that may run to several pages. Each additional document improves the effectiveness of the modules trained in this way. Making the data available in digital form, e.g. in structured databases or Excel, ensures it can be reused.
The specialists in the IT department identified a need for action to ensure operational capability in coordination with the company-wide cost-cutting measures and to respond to the growing application landscape of RWE Power AG. Their objective was to relieve the available capacities of the labour-intensive task of maintaining source code as much as possible in order to continue maintaining and adapting the applications developed in-house into the future.
To avoid having to write standard code for different digital applications over and over by hand, or to import it using copy-and-paste, the team in the IT department at RWE Power, the “Power Core IT”, switched to having this code created using “code generators” developed in-house.
A source code generator is a software program that generates the source code using a selected programming language based on models (e.g. UML), another formal language (e.g. XML), or other formats (e.g. decision matrices). Domain-specific program code is generated in this way, based on available meta-information.
These source generators can automatically generate recurring programming patterns with no time cost for the developer.
Depending on where the software module is used, this could account for up to 80 percent of the source code, which could mean hundreds of thousands of code lines depending on the size of the program. The generators also shift the development process to a certain extent, away from programming and towards modelling. Maintaining models is often easier than maintaining source code, since models enable a faster overview and are more independent of the programming style of an individual developer.
Models allow formal checks for compliance with established restrictions, specifications and other requirements.
Code generators are also not at risk of careless errors. They generate code that automatically complies with the programming guidelines. The generated code is consistent, and its quality is assured by the generator.
The constant refinement of programming languages necessitates regular maintenance projects.
These ensure the applications are compatible with new languages or language versions, and that they represent the state of the art. With generated code, “only” adaptation of the code generator is required, not adaptation of the entire applications.
With this approach, RWE Power is able to concentrate entirely on implementing the business logic and the technical processes, and thus achieving the maximum possible added value based on the available expert knowledge. This improvement in efficiency and productivity in developing applications means a further element is in place to provide valid support for the application landscape at RWE Power AG.
