by Cliò E. Agrapidis, PhD
As industries push for greater efficiency and sustainability, automated control systems are becoming essential in managing complex operations. From industrial processes to smart home devices, we can delegate decisions and control to a digital companion. But how are they able to decide what to do?
How Automated Systems Make Decisions
To give an example without considering the latest technological devices, we can think of a humble thermostat. As the days get colder, we set up our heating system through a thermostat. Usually, this means programming a daily or weekly schedule, setting the desired temperature at specific times—typically higher during the day and lower at night. Once the program is set and the heating is turned on, we no longer need to manually switch it on and off to maintain the desired temperature. Through temperature sensors, the thermostat takes care of activating the heater when needed.
If we leave our homes for a vacation, we usually manually switch the thermostat to “vacation mode,” lowering the desired temperature to avoid energy waste (and higher energy bills). But how could this technology make that decision without manual input? If we include additional sensors—such as a carbon dioxide sensor, window open-close detectors, or outdoor temperature monitors—the thermostat system could collect more data, store it, and analyze it. This would allow the thermostat to adjust its settings without human intervention.
But there’s more this system could do: based on its data analysis, it could predict when you are coming home when you go to sleep, and when you wake up, adjusting its settings accordingly. In this way, we would have created a digital twin of our heating system. In industrial settings, digital twins operate in a similar way—continuously analyzing data to optimize system performance in real time.
The SUSHEAT Digital Twin: Optimizing Industrial Heat Systems
The same principle of prediction and optimization applies to far more complex systems, such as those used in industrial heat generation. In developing a new technology for this purpose, the SUSHEAT project includes the creation of a digital twin for optimizing the efficiency, flexibility, and reliability of industrial heat supply, integrating high-temperature heat pumps, thermal energy storage, and solar concentrators. The concept for this digital twin was outlined in a recent publication, part of the conference proceedings of the International Conference on Industry Sciences and Computer Science Innovation 2024.
As the SUSHEAT researchers explain, “the concept of a digital twin involves the creation of a precise and intricate replica of a physical object or system.” This is no easy feat when it comes to complex systems such as the SUSHEAT high-temperature heat pump.
In order to create a replica, it is necessary to have a clear scheme of the system being mirrored. The SUSHEAT technology can be broken down into its core components and their respective interactions. Importantly, the system makes use of renewable energy sources, which are neither reliable nor constant over time. Natural, unpredictable factors introduce chaotic and probabilistic elements that are difficult to model. To overcome the reliability problem, engineers can integrate multiple energy sources, creating a more resilient but complex system.
The Role of AI and Data in Managing Renewable Energy and Beyond
Traditionally, operating such systems relied solely on mathematical models or schematics, which require extensive expertise in the field. SUSHEAT suggests a new solution: having “a digital twin connected in real-time to the high-temperature heat upgrade system, allowing for a significantly more seamless and optimized operating experience.”
The authors outline an architecture for the digital twin, defining which sensors (to gather information) and actuators (to perform tasks) are needed, while also emphasizing the role of AI in shaping the digital control hub. Human interaction is incorporated through the development of a user interface, allowing on-site or remote control via the Cloud, where experts can access historical or current data and make informed decisions faster and with greater accuracy
A Step Toward a Sustainable Future
The complexity of this system requires additional steps to refine the architecture and define the most suitable AI and machine learning methods for the available data. SUSHEAT researchers are working within a multidisciplinary team to create a digital twin. To facilitate the transition toward a low-carbon industrial heat supply. The final goal? Managing the unpredictability of renewable energy sources and making them more reliable to promote widespread adoption, helping achieve the EU’s net-zero goals by 2050.
