Artificial Intelligence for Steam

17 Dec 2019

Reading time: 2 minutes

Applying AI to industrial boilers allows to increase efficiency, improve self-diagnosis, to achieve the predictive control of the steam generator and its complete integration within the production process.

How can the Artificial Intelligence combine with a machine of the end of ‘700?

Let’s start with the definition of artificial intelligence: it is a particularly complex discipline that studies whether and how it is possible to artificially reproduce complex mental processes.

Moving this concept to boilers means processing and applying algorithms and mathematical models able to solve some classic steam generator problems, such as the typical inefficiencies of transient conditions and the instabilities due to discontinuous processes or bad setting of control parameters (human error). But how can these models be created?

Thanks to today’s PLC computing power, extensive sensors and special software, a conspicuous series of data is collected, stored and processed, used to automatically optimize the process and make it more efficient.

Efficiency that translates for customers into savings.

A smarter steam generator

In order to make its generators more and more intelligent, Cannon Bono Energia has developed a monitoring system based on a cloud platform, which allows to collect and store data relating to the steam generator, to process them and made them available on multiple devices. Using a cloud allows to integrate the boiler with the entire production process, with which exchange data.

Therefore, data are not only available through a system – OptiCloud – which helps the customer to always keep his boiler under control, being able to view all the parameters in real time through customized dashboards; but these data also feed an MPC (Model Predictive Control) created by Cannon Bono Energia in collaboration with Politecnico di Milano, through which the boiler is able to modify its operation based on the effects that this modification will cause in the future moments according to a goal.

The MPC thus allows the boiler to operate more homogeneously, improving transients, avoiding swings that can wear out the different components and, above all, allows the generator to work always around the optimum.

This translates into a more precise control and an improvement in the overall efficiency of the boiler.

That is: savings.

In this context, however, everything starts from the data – concerning the machine and the whole process: without data it is not possible to create models and, therefore, to apply artificial intelligence systems to the boiler.

In summary, what does it mean having a boiler with Artificial Intelligence?

It means having a machine that can operate always at the optimum level, even when boundary conditions vary. It means improving control performance and working efficiently. It means saving.