Category: Technology

The Digital Twin at the Center of Renewable Energy

According to the German Wind Energy Association (BWE), the share of wind energy in German electricity production this year is 27 percent, and in 2020 wind energy even represented the most important energy source in the German electricity mix. In total, more than 31,000 turbines have been installed, saving 89 million tons of CO2 equivalent in 2019. Wind power is thus a mainstay of low-CO2 and sustainable energy generation and makes an important contribution to the energy transition. Further increasing yields while reducing maintenance costs is therefore of great importance.

Increasing the efficiency of wind farms with smart systems

Digital Twins are the central element in exploiting the full potential of wind power and maximizing yields. Driven by the vision of creating a data-based development tool for the wind industry, the WIND IO joint project, funded by the German Federal Ministry for Economic Affairs and Energy, started a year and a half ago.

Under the leadership of the Institute for Integrated Product Development BIK at the University of Bremen, we are working with several consortium partners to build research facilities as cyber-physical systems and retrofit them with sensors, electronics and computers known as IoT gateways. This makes it possible to digitally map all the operating information of the real plant and combine it on a digital twin. The operating behavior can be simulated on the basis of the Digital Twin, which in turn provides insights for further optimization of the wind turbine. The Digital Twin not only provides information about the current energy yield, but also offers a comprehensive overall picture of the condition of each individual turbine.

Improved installation, maintenance and overhaul processes

The information obtained can be used, for example, to optimize maintenance and overhaul processes. For example, the data makes the aging process of components transparent at all times and automatically triggers an alarm if defined limit parameters are exceeded. The Digital Twin also uses the operating, environmental and weather data collected to determine a favorable time for maintenance of the plant. Ideally, this should be carried out when there is little wind, so as not to be at the expense of energy generation.

Both statistical methods and Artificial Intelligence (AI) models are used for the calculations. These methods also help to determine the best time to assemble a wind turbine, since the rotor blades can only be installed under certain conditions. For this purpose, in addition to weather data, additional parameters such as the vibration of the tower are included in the calculations.

Digital Twins for a sustainable industry 

The WIND IO project vividly demonstrates the potential of digitization and especially the concept of the Digital Twin. In addition, companies can use their data to simulate entire production and operating cycles. This makes it possible to minimize resource consumption, reduce energy consumption and at the same time coordinate production steps more effectively and optimize transport routes. Concepts such as the Digital Twin and data-intensive analysis methods are thus essential for a gentle and efficient industry.

Successful IoT business: just a question of standards?

There are days the little things in life make me happy. When my microwave broke last week and even a repair couldn’t save it, it took me less than five minutes to solve the problem: simply selected a new model on the manufacturer’s site using my smartphone, ordered it and paid via PayPal. Three days later it was unpacked, plugged in and running. The ease of this process illustrates two things:

  1. digitization makes it incredibly easy for us to handle even extensive processes quickly.
  2. I didn’t ask myself whether the microwave would also fit into my power socket and whether it would meet the usual standards for radio interference suppression, hazardous substances, etc.

Anyone who has ever traveled abroad knows that this lack of concern is not a matter of course. In the case of power sockets, the right time was simply missed to ensure global standards. In the meantime, the implementation of a standard would cause so much cost and electrical waste that it is no longer practicable.

Unimaginable that something like this could happen again to our highly developed society… or could it?

Digitization is opening up new business potential. The focus is shifting from the exchange of physical goods to the exchange of information. When I buy my microwave, it’s not just the manufacturer who earns money, but also the online payment service PayPal. And that is solely through the exchange of information. Digitization is also creating the basis for new business models in industrial companies. This is shown by a recent study by Sopra Steria and the F.A.Z. Institute. More and more machines and systems are being networked via IoT platforms in the industrial Internet of Things in order to determine performance data or offer product-related services. This is a development that has taken hold around the globe and is thus giving rise to many solutions with different data models and integration options. This allows us to draw a worrying parallel to the connector mess mentioned above. Companies that want to drive their digital business forward quickly lose their orientation here when choosing an IoT solution that is suitable for them. After all, how future-proof it is depends largely on how well it can be connected to other systems and data sources.

Global standards for sustainable digitization

Serious initiatives here give hope for an international standard in the industrial Internet of Things. The Plattform Industrie 4.0, for example, has developed the concept of the management shell, which is to be understood as the digital representation of a device. It makes it possible to address machines with all the necessary information and functions. For example, I could develop an app for my microwave, interact with it, display the instructions for use, and set the power intensity or duration via smartphone. If the manufacturer of my washing machine also provides the information and functions of this device according to the management shell concept, it is no effort for app developers to integrate other devices into their application. This manufacturer- and system-independent interoperability paves the way for the future of Industry 4.0.

AI – Where we are in the Hype Cycle and how it continues

While the artificial intelligence index shows that the increase of research articles and conferences in the field of AI continues, the media is slowly showing some fatigue in the face of the hype. So it’s time to take stock: What has been achieved? What is practically possible? And what is the way forward?

What has been achieved?

In the years 2018 and 2019 the previously developed methods for the application of neural networks (this is how I define AI here) were further refined and perfected. Whereas the focus was initially (2012-2016, Imagenet competition) on methods for image classification and processing and then on audio methods (2015-2017, launch of Alexa and other language assistants), major advances in text processing and generation were made in 2019 (NLP = natural language processing). Overall, the available technologies have been further improved and combined with a great deal of effort, especially from the major players (Google, Facebook, OpenAI, Microsoft).

What is practically possible?

The use of AI is still essentially limited to four areas of application:

  • Images: image recognition and segmentation
  • Audio: Conversion from speech to text and vice versa
  • NLP: word processing and generation
  • Labeled Data: Prediction of the label (e.g. price) from a set of features

This list is surprisingly short, measured by the attention AI receives in the media. The most impressive successes of AI, however, result from a combination of techniques such as speech assistants using a combination of audio, NLP and labeled data to convert the input into text, recognition of text intention with NLP and prediction of the speaker’s wish by using huge amounts of labeled data, meaning previous evaluations of similar utterances.

Decisive for the development of precisely these AI application fields were

  1. the existence of large quantities of freely available benchmark data sets (data sets for machine learning) on which algorithms have been developed and compared
  2. a large community of researchers who have jointly agreed on the benchmark data sets and compared their algorithms in public competitions (GLUE, Benchmarks AI, Machine Translation, etc.)
  3. a free availability of the developed models, which serve as a starting point for the practical application (exemplary Tensorflow Hub)

Based on these prerequisites one can quickly assess how realistic some marketing fantasies are. For example, there are neither benchmark data sets nor a community of researchers for the often strikingly presented field of application of predictive maintenance, and accordingly there are no models.

What’s next?

On the one hand, it is foreseeable that the further development in the AI area will certainly continue initially in the above-mentioned fields of application and continue to develop in the peripheral areas. On the other hand, areas are emerging which, similar to the above-mentioned fields of application, will be driven forward at the expense of large public and private funds (e.g. OpenAI and Deepmind are being subsidised by Elon Musk and Google with billions of euros respectively). An example of large investments in this area is certainly autonomous driving, but also the area of IoT. In total, I see the following areas developing strongly in 2020-2022:

  • The combination of reinforcement learning with AI areas for faster learning of models
  • A further strengthening in the area of autonomous driving resulting from the application and combination of AI and reinforcement learning
  • Breakthroughs in the generalization of the knowledge gained from image processing to 3D (Geometric Deep Learning and Graph Networks)
  • A fusion of traditional methods from statistics with neural networks
  • IoT time series (see below)

I see a big change coming with the rise of IoT and the associated sensor technology and data. By their very nature, IoT data are time series that must be filtered, combined, smoothed and enriched for evaluation. Relatively little specific has been done to date for this purpose. It could be that from 2020 – 2022, this topic could hold some surprising twists and breakthroughs for us. German industry in particular, which has benefited rather little from the initial developments in the field of AI, should find a promising area of application here.