Taking sustainable production to the next level

Part 2: Research project "PHI-Factory"

2021/03/30 by

If industry wants to face up to the challenges of the energy transition, it must rethink production – a complex transformation process that TU Darmstadt supports with new digital and AI-based solutions. The scientists behind it are application-oriented, interdisciplinary and holistic. Their common goal: to promote emission reductions, resource conservation and flexible energy use technologically. And in doing so, also to contribute to more companies putting environmentally and economically sustainable production on their strategic agendas. In the ETA Factory, researchers and their practical partners have a model factory on the Lichtwiese campus where they can validate their innovations and make them ready for practice.

Flexibility in energy management

They are working on the energy management of tomorrow. In addition to energy efficiency, researchers at TU Darmstadt also rely on synergies between the sectors and connection to the energy market.

The supply technology of an industrial plant consists of a variety of technical systems. Aside from the actual production chain, energy converters and storage systems must also be orchestrated in a complex system. Simple monitoring systems are commonly used for this purpose, which primarily document electrical consumptions. “This provides a rough overview of the plant's total energy consumptions; it is not possible to assign consumption to individual plants, machines or products with these systems,” says Professor Matthias Weigold of the Institute of Production Management, Technology and Machine Tools (PTW) at TU Darmstadt. However, at a time when volatility is increasing along with the upturn in electricity generation from renewable sources such as wind and solar, far more flexible energy management systems are required as well as more efficient production facilities.

“There is much potential for more efficiency and flexibility in industrial manufacturing. We just need to raise it,” emphasises Professor Stephan Rinderknecht, Head of the Institute of Mechatronic Systems (IMS). A look at the ETA Factory shows how this can work. Within the project “PHI Factory”, a framework has been developed and tested on a real scale with which industrial companies can save costs and resources and also support the electricity grid with flexible energy management.

There is much potential for more efficiency and flexibility in industrial manufacturing. We just need to raise it.

Development of new model factory

In cooperation with companies such as Bosch Rexroth and Software AG, the PHI Factory team has turned the ETA Factory into a fully digitised and energy-flexible model factory in which all energy and process data is monitored. “This also allows a prediction of energy consumption as well as providing greater transparency – like a weather forecast,” reports Weigold. Predictions of when the next peak load will occur are possible based on the system with a lead time of between 15 minutes and one hour. The forecast horizon for a single machine is only a few minutes. Coupled with weather and market data, production can be planned in an energy-optimised manner and the control mode of supply technology can be adapted flexibly and quickly to changing conditions.

One important component in the new energy management system is a new hybrid and highly efficient storage system that combines a flywheel storage system with a lithium-ion battery, taking advantage of both technologies. “If I want to store larger amounts of energy for a longer period of time I use the battery, and if lots of machines reach the load peak at the same time I use the other storage because it releases the energy quickly,” explains Georg Franke, research associate at the IMS. This means that the factory only uses the grid when it needs electricity, and can operate autonomously for one day, for example if there are low feed-ins from renewable energy sources.

As well as the production plant, the electrically powered factory vehicles are also integrated in the system as a further consumer. Load peaks can be shifted when the vehicles are parked for charging, i.e. the charging process only starts when electricity is cheap or is no longer needed elsewhere. In the next step, the batteries of the electric vehicles are operated bidirectionally, i.e. they are also able to release energy into the network at the charging station. “This is a good example of how we will be able to exploit synergies between industry and the mobility sector in the future,” says Rinderknecht. “Without innovative solutions such as these for sector coupling, neither the energy nor the transport transition will be feasible in the long term.”

Furthermore, the flexible factory of the future will contribute to stable load management. “The system always knows how the network is doing at a particular moment,” says Franke. This not only promotes efficiency and sustainability but also offers economic benefits, for example when a production plant generates revenue by purchasing the surplus in the event of an oversupply from renewable energies. “So it is important that we always include economic issues as well as technology in our scientific considerations,” emphasises Rinderknecht.

Involvement of artificial intelligence in energy management

Without methods of artificial intelligence such as machine learning, which was already used in the context of “PHI Factory”, it is not possible to achieve a holistic, company-wide, cross-sector energy management system that is adapted to the market. Within the “KI4ETA” project, which is just starting, solutions are now being sought to manage real-time data streams automatically and to operate production and supply in a factory with the use of AI algorithms, from the collection of data to the control of each individual machine.

“We want to provide energy managers and consultants and those responsible for production and infrastructure with tools to analyse complex factory systems and create low-carbon emissions,” Weigold states. An energy management platform that is easy to install and operate and can be further developed towards cloud and edge computing, the energetic coupling of entire sectors, automated transparent monitoring and AI solutions for energy optimisation: the new research project has much potential for springboard innovations.

Facts and figures:

The joint project “PHI Factory” (duration 12/2016 to 03/2020), in which the Institute for Production Management, Technology and Machine Tools (PTW), the Institute of Mechatronic Systems (IMS) and the Department of Electric Energy Supply using Renewable Energies (E5) of TU Darmstadt participated as well as the consortium partners of the ETA Factory, was supported by the Federal Ministry of Economics. The “KI4ETA” consortium, led by the PTW and in cooperation with the IMS and other industry and research partners is expected to launch in April 2021. Researchers at the PTW and the ETA Factory have also been contributing their expertise to the Kopernikus project “SynEergie” funded by the Federal Ministry of Education and Research (BMBF) since 2016.

More about sustainable production

“On the way to an intelligent productt”: Read also the first part on sustainable production at TU Darmstadt.