Verbund testing different AI technologies for both generation and retail use cases

Last update on March 16, 2020.

Verbund testing different AI technologies for both generation and retail use cases

Austria´s leading utility Verbund has created a new digital execution department to help different business units with digitalisation projects.  

The team is already examining which technologies are the best fit for various challenges in Verbund’s operations, from its more than 100 power plants to the retail supply group. Verbund has a demonstration power plant in Rabenstein, Styria under the Digital Hydropower Plant 4.0 project in conjunction with the European power plant association, the Graz University of Technology and with technology partners.  

It has a diving robot that can inspect the turbine and a digital twin to help detect anomalies and allow predictive maintenance.  One of the first tasks of the new digital execution team is to streamline the data flowing from the power plant and from the Internet of Things components to ensure it is the right quality and at the right timescale. Many of Verbund’s power plants are already highly automated so the basis for data gathering is good, but the challenge will be in centralising and streamlining data from all the power plants together, says product manager Hans-Linus Pfau.  

Verbund took the decision to hire a team of physicists who can also code, as problem solving in utilities needs both worlds combined, he says. “You can be way more agile and efficient if you can combine the two, building on your own expertise and testing new technologies,” he says. “We have to evolve into more of a software company. It’s a big change for utilities.” 

Artificial intelligence can be used for anomaly detection in machine turbines, but Pfau’s team is still testing whether robust regression model or neural networks work best for that application. Robust regression is a more established technology compared to neural networks, he says.  

The team is also researching which technologies should be applied in the retail sector to help customers who want to switch supplier. Image recognition can facilitate the process of gathering the data that is required to onboard a new customer so that they only have to send in a previous bill instead of searching for the relevant information themselves. For commercial and industrial customers, Verbund is developing a way of using historic consumption data and production plans to determine peaks in order to be able to give the customer feedback to offer peak shaving opportunities, sometimes in conjunction with a battery system.  

Verbund is working on hydrogen projects, including the H2FUTURE project to produce green hydrogen for the steel industry, as well as researching how to optimise electrolysers to produce power from hydrogen. A prototype is already going through internal testing and will enter a pilot phase later this year.  

Verbund is also involved in a European mobility programme called SYNERG-E looking at asset-based services including lithium-ion batteries. It aims to reduce the load on the power grid, store renewable energy for electromobility and provide grid-related services. SYNERG-E is co-financed by the Connecting Europe Facility of the European Union.

Next entry

Previous entry

Similar entries

Comments

No comments yet.

Pingbacks

Pingbacks are open.

Trackbacks

  1. Slot game pc on 07/07/2020 6:29 a.m. #

    Smartgrid Ireland - Verbund testing different AI technologies for both generation and retail use cases

  2. sbo euro on 07/07/2020 6:32 a.m. #

    Smartgrid Ireland - Verbund testing different AI technologies for both generation and retail use cases

  3. gta online: the diamond casino & resort on 07/07/2020 6:34 a.m. #

    Smartgrid Ireland - Verbund testing different AI technologies for both generation and retail use cases

Trackback URL

Post your comment