“By their very nature, heuristic shortcuts will produce biases, and this is true for humans and artificial intelligence, but the heuristic of artificial intelligence is not necessarily the same as that of humans.” Daniel Kahneman, Nobel Prize in Economic Sciences (2002).

The technological achievement of industrial process automation was emblematic for society and opened new transformative perspectives from the change in the operational conditions of production processes automatically, without human intervention. In addition to concrete economic and operational gains, the automation of operations represents an unprecedented opportunity for integration.

This integration brings together different levels and areas of activity of the companies, which gain synergy and agility in decisions, including the supply chain, agricultural and industrial operations, up to the end customer. The automation journey begins in the second half of the 18th century, with simple and semi-automatic devices, indicating that it was possible and more efficient to control operational variables automatically.

With the technological evolution in the industries, pneumatic devices made it possible to control one variable by changing another: controlling a temperature, for example, opening or closing a valve based on simple rules of proportionality between variables. Any initiative that ventured beyond single-variable control with proportionality to the error of the controlled variable would already be considered advanced control.

Control devices evolved into sophisticated computers and numerous possibilities for calculations and control algorithms became available, including virtual sensors that complement and even replace traditional physical sensors. Next-generation process automation already operates multivariate and predictive systems as established technologies.

Recently, in the sugar-energy sector, industrial units have emerged with a high degree of automation, with a centralized Industrial Operations Center and cogeneration plants that integrate backpressure turbines with other condensation turbines and high pressure boilers, enabling the implementation of advanced optimization and control techniques.

The sugarcane biorefinery favors the search for optimization in real time, with advanced control algorithms in the search for greater operational stability and better efficiencies. The implementation of an effective industrial automation program must consider some premises and recommendations that can guarantee the success of the application of the digital technologies of Industry 4.0 in the context of the sugar-energy sector.

a) Operational flexibility: Each solution adopted must be able to establish simple and secure communication (cybersecurity) with other systems and sensors in general. In practice, it is necessary to keep systems with open architecture and available for aggregation of new control, simulation and optimization tools and technologies, according to business needs.

b) Modularity: In synergy with operational flexibility, the automation and control should preferably be modular. The possibility of modular use of the control technology, by areas or by specific sections, stimulates the continuous improvement of industrial performance in a sustainable way, according to objective evaluations of the results and consolidation of captures at each stage.

c) Operational management: Consists of monitoring indicators of interest (Key Performance Indicators) for the definition of operational guidelines and for anticipating the main trends in the process.

The predictive capacity with anticipation of trends in the production process is, perhaps, the main contribution of automation and artificial intelligence. In addition to the balances made by the industrial laboratory of the plants, the increasing application of tools for in-line control (monitoring and control of the process carried out by instruments placed directly in the production process, “in line” with the production flow) (on line) in industrial processes must change the way plants operate.

In-line control with NIRS (near infrared spectroscopy) probes and other advanced analytical techniques can significantly contribute to the operational stability of industries, with improved yields and efficiencies. Along with in-line control, self-control initiatives (monitoring and analysis of process variables carried out by the operation teams, with the support and support of the quality and laboratory teams) ensure speed and complete engagement of the operation in the operational optimization of the industrial units.

The configuration of an advanced automation system also requires some heuristic analysis to provide values to be loaded in the controllers and guarantee both performance and operational safety. Furthermore, the robustness of the system can be ensured through its ability to react to some instrument error, through data reconciliation or verification techniques, such as adaptive neural networks. In all situations, it is essential to maintain qualified technical teams to ensure excellence in the tuning of meshes and monitoring of results.

In practice, the application of Artificial Intelligence to industrial processes brings concrete perspectives of gains. This trend is confirmed, for example, by the increasing robotic density, which already reaches a world average above 125 robots per 10,000 workers. In Brazil, we are still well below the world average and far from leading countries such as South Korea, Singapore, Japan and Germany. Brazil's ability to evolve and integrate new technologies efficiently will ensure its long-term competitiveness.

But technological choices are not trivial decisions. On the contrary, they represent risk management exercises in the search for operational excellence and better results, in a broad and integrated view of the business in the long term. Not adhering to new technologies is also a choice and can have irreversible consequences; therefore, we must always be attentive and monitor the opportunities offered by the evolution of technical knowledge. In the sugar-energy sector, industrial maintenance is likely to be the next wave to accelerate the application of Artificial Intelligence in industries.

The importance of industrial maintenance for the business, especially in cost management and industrial availability, justifies its role in the technological evolution for asset monitoring and predictive maintenance. With the application of concepts of precision mechanics and artificial intelligence to ensure maintenance routes, the solutions will be increasingly integrated, fast and agile, as is already happening in leading companies in the sector.

This evolution is continuous; we are destined to participate in ecosystems that will incorporate new technologies with increasing speed. Although it seems paradoxical, technological transitions require thoughtfulness and boldness at the same time, and each country, each industry and each sector must build its own strategies. We just cannot be oblivious to the movements that are taking place.