Decarbonizing industry thanks to AI: myth or reality?

Artificial intelligence (AI) is gradually becoming a key technology in many fields, and energy efficiency is no exception. By integrating intelligent systems that can analyze, predict, and optimize energy consumption, AI could well transform the way we use and manage energy.

Through the example of the words of the founder of Purecontrol, Gautier Avril, during his speech in the Safety Innov podcast, we are going to explore how AI is already contributing to this energy revolution.

‍

1 - Make technology one of the levers for the decarbonization of industry

The global energy system is undergoing a real transformation, especially since the recent energy crises. With much higher and much more volatile prices, manufacturers may be tempted to deprioritize their programs to electrify uses. For all that, there is an urgent need to decarbonize, starting with the search for energy efficiency.

A tool for performance analysis, decision support, or even real-time optimization, this innovation is reshaping the world of energy. Gautier Avril explains that he became aware of the challenges of global warming about fifteen years ago. His company, Purecontrol, is committed to using technologies, especially AI, to contribute to the fight against global warming.

“Using technology to fight global warming is not going to solve everything, but AI is still an opportunity to significantly improve certain aspects of our energy management.” adds Gautier Avril‍

Many solutions have emerged in recent years, ranging from simple performance analysis to real-time optimization.

‍

2 - Exploit existing data thanks to digital twins

Digital twins are virtual models of physical installations that allow you to simulate their operation and test optimization scenarios. In short, a digital duplicate of industrial site processes is created, allowing access to dashboards that show accurate data. This in-depth data analysis makes it possible to better understand the relationships between different parameters, and to create accurate predictive models. The information provided allows site managers to adjust their operations according to the observations made: energy drifts, excessive production quality, equipment failure, etc.

Many companies have specialized in creating and exploiting digital twins for analysis and optimization purposes. Purecontrol is one of them: by connecting to industrial automatons, the solution retrieves operating data from factories or water management sites.

” The sensors present on these installations are often only operated at 10% of their real capacity. AI makes it possible to retrieve all this data, add external information such as weather forecasts, and analyze it continuously using machine learning algorithms.”

Digital twins are a powerful asset in understanding and making processes more efficient. However, they have their limits: adjusting industrial mechanisms requires the continuous involvement of operators in the process. And often, the coding of automatons is set in stone, lacking flexibility and reactivity in the face of the evolution of various parameters (product quality/treatment, quantity of products used according to the factory load, etc.).

While digital twins are effective in identifying savings opportunities, their application can sometimes be challenging. Some savings opportunities have been known for a long time, but their implementation is often too complex.

” For example, on a site equipped with solar production, it seems obvious to want to consume energy when there is sunlight. But the number of parameters to be taken into account is too important: the vagaries of the weather, the quality of the forecasts, the inertia of the process, etc. In the best case, we will simply postpone certain uses at noon, hoping for the weather to be nice. The tools capable of finding effective management strategies, capable of adapting to these hazards, are still too few.”

‍

3 - Go further in optimization with real-time control (control-command)

Using AI to find these optimization strategies is at the heart of the creation of Purecontrol. For Gautier Avril, the trigger came from the victory of AlphaGo, the first success of a machine against the Go champions, which showed that artificial intelligence was capable of finding new strategies to solve complex problems.

”By transposing this technology from the gaming world to real industrial problems, artificial intelligence has great potential for Help energy-consuming manufacturers achieve their goals of sobriety while respecting business constraints.” 

Purecontrol's innovation, and the strength of the solution, is to putput AI at the heart of “control and command” The solution consists in letting the AI find management strategies and directly apply the recommendations to automatically control industrial processes, without an intermediate algorithm. All that remains is to give the rules of the game: respect operating constraints while optimizing energy consumption and minimizing greenhouse gas emissions. The idea is to go beyond simple data analysis and the construction of recommendations, to implement dynamic management in real time.

”While the best co-driver will never allow an average driver to become a rally champion, the best models will never allow an automaton program to take into account all the subtleties of a complex environment.”

Concretely, the Purecontrol management solution continuously sends commands, which implement optimization actions, defined by AI algorithms. This management consists of adjust the processes in real time according to the evolution of the parameters, while meeting the given objectives and constraints. For example, in a wastewater treatment plant, if the objective is to reduce treatment chemicals, the solution will automatically adjust the injection of these products according to the quality of the incoming water and the weather, which has an influence on water treatment. This principle makes it possible to limit the quantity of chemical products used while ensuring compliance with environmental standards.

‍

4 - Extending applications to achieve energy efficiency

By analysing the impact of each decision, Algorithms are constantly adapting and improving. This flexibility is valuable in particularly complex and changing environments. This is why the Purecontrol solution significantly improves the functioning of many use cases: thermal regulations subject to strong hazards (incinerators, galvanizing), biological and chemical treatments, which are very difficult to control (treatment plants) or mesh networks where decisions on each node will have consequences on the entire network (wastewater networks, heating networks, etc.).

”Earnings vary depending on the sites and technologies used; but they can vary from 10% to 40%. Even better, with the Purecontrol solution these results are obtained without installing new hardware, but simply by optimizing the use of existing data.”

The example of Purecontrol is a perfect example of how AI can revolutionize the industrial world. By making full use of available data and using advanced algorithms to optimize processes, AI offers innovative and effective solutions to reduce our energy consumption and minimize our environmental impact while stabilizing the operation of the installations. This technology is still evolving, and its applications are constantly expanding.

”We are able to use data more intelligently, to predict and optimize energy consumption, and to reduce our carbon footprint significantly. New energy-saving opportunities will emerge as AI continues to develop and integrate into new areas.”

‍

By embracing these innovations in the right way, we can look forward to a more sustainable and environmentally friendly future.