Reapplication of the MM Project (Monitoring Mills) in a sugar, ethanol and other products production unit aiming at the validation of an artificial intelligence algorithm

• Grant number: 22/11309-1
• Support Opportunities: Research Grants - Innovative Research in Small Business - PIPE
• Start date: December 01, 2022
• End date: November 30, 2023
• Field of knowledge: Interdisciplinary Subjects
• Agreement: FINEP - PIPE/PAPPE Grant
• Principal Investigator: Wokimar Teixeira Garcia
• Grantee: Wokimar Teixeira Garcia
• Company: ITC Consultoria em Engenharia Ltda. - ME
• CNAE: Fabricação de açúcar em bruto ; Fabricação de álcool
• City: Piracicaba
• Pesquisadores principais: Jaime Finguerut ; Joaquim Henrique da Cunha Filho ; Sergio Suzuki

Abstract

Instituto Tecnologia Canavieira (ITC) is a company whose mission is to generate value by bringing innovations and tools that will allow better production management for the sugar-energy sector with engineering, chemical and information technology solutions. It considered as a window of opportunity to develop a monitoring tool, based on data provided by NIR sensors (Near Infra Red) in real time, unprecedented in the market, which aims to provide managers with information to improve the efficiency of the production process. Based on the concepts of I4.0, it evaluates the performance of the fermentation process and sugarcane milling operation, and is an Analytical Intelligence tool, "Project MM - Monitoring Mills", which uses sensing technologies (NIR) for the online monitoring of analytical data on the composition of raw materials and the main streams responsible for most of the losses, in addition to analyzing the outputs of processes and operations, which will allow plants to make safe decisions in a faster response speed for analysis, and ensuring supervision, understanding and implementation of control strategies that will have a positive impact on industrial efficiency. Considering that the company is dedicated to the provision of engineering services, it needs to establish partnerships with companies or entities for the execution of each of its development projects aimed at carrying out one of the activities related to the methodology that the ITC employs. Thus, in 2021, the pilot project began, using its own resources and that of a commercial partner, Usina Açucareira São Manoel S/A, and it is in this unit that the "MM Project - Monitoring Mills" is being conducted. It was established that the plant and the proposing company would indicate the operations (milling and fermentation) and equipment that would be monitored, as well as the commitment to provide infrastructure and historical data (already existing) for the conduction of the project. In the first phase of the project, the technological and application concepts of this project were formulated, defining the critical process in an experimental way and the proof of concept in the unit. The company submitted, together with Phase 2 of FAPESP's PIPE Program, a project aimed at continuing the pilot project, "Project MM - Monitoring Mills", with the objective of raising online profiles of the composition of the prepared cane, bagasse, must and centrifuged wine; create a database, by capturing data online and from other sources, in addition to historical data to then use exploratory statistical analysis techniques, seeking to confirm and quantify the relationship between the possible causes and the main losses quantified by the sensors; develop and implement conventional control loops using sensors that have already been calibrated and audited, allowing for more efficient control, whether in terms of strategy or speed of adaptation to changes; and carry out the evolution of the database, to a machine learning system, capable of recognizing the current situation and comparing it with an "ideal" situation, based on the history and the knowledge of the "first principles" (knowledge of the capabilities of the equipment, mass and energy balances). This system will gradually interact with operators and supervisors with messages and proposition of actions ("open loop") to then assume some controls in "closed loops", that is, the system itself assuming set-point changes for example. The submission of a proposal in Phase 3 would represent resources to complement activities in Phase 2, seeking to validate the Artificial Intelligence algorithm, with the implementation of the project in a second sugar-energy unit, in order to aggregate process data from a second production unit, which eventually may bring new characteristics of the raw material and the plant's production process, and, in order to start the commercial development of this ITC service. (AU)