Introduce tu búsqueda a continuación y presiona el botón para obtener resultados.
Juan de la Cierva: Efficient and circular industry
Real-time modeling tools for manufacturing processes including energy flows to improve product quality and circularity
Program: JUAN DE LA CIERVA-INCORPORATION GRANTS, 2020 CALL FOR PROPOSALS
Reference: IJC2020-043717-I
RESOLUTION Date: 11/29/2021
Date of execution: 01/05/2022-30/04/2025
Manufacturing systems are complex because many parameters collectively influence the energy performance of production processes, including those related to the environment, components, use of materials, machines, cells, lines and supply chains. Improving industrial energy efficiency generally requires the integration of energy data into production management systems, such as historical data, real-time data as well as real-time forecasted energy costs. All this information is generated by collecting sensor data and generating additional information through relatively simple models based, among other techniques, on multi-agent systems. However, the detailed performance of machines, processes and final product (including their quality and expected performance in use) is fundamental to the evaluation and is usually not adequately considered and is incorporated through empirical algebraic models. In this framework and taking into account a growing awareness of circularity issues, it is necessary to develop novel approaches that support, based on detailed, complete and instantaneous knowledge, the improvement of life cycle cost analysis in energy-intensive manufacturing companies.
To develop and implement novel approaches to improve life cycle cost assessment, there are three main barriers: (1) the popularization of improved computational tools for energy assessment that provide rapid access to valuable information (knowledge) and allow easy implementation; (2) the horizontal integration of knowledge along the entire value chain; (3) the vertical integration of knowledge and its contribution to lean manufacturing approaches.
To meet the challenges described above, we propose the use of a combination of ICTs to move from diagnosis to prognosis in the control of life cycle cost assessment in manufacturing, while preserving product quality and confidentiality.
Objective 1. Enable the analysis of energy flows with thermodynamic information. Despite the recent popularity of machine learning in the industry, the development of efficient models for energy data analysis remains a challenge. We propose to adapt existing machine learning libraries developed at ITAINNOVA to allow the calculation of integral balances from available data. In this way, data analysis becomes physically consistent, enabling robust modeling approaches. In addition, this method will be integrated with real-time simulation information, calculated by CFD or FEM, of the manufacturing processes.
Objective 2. Integrate in real time the available vertical and horizontal knowledge. The same data structure used for the machine learning algorithms developed for Objective 1 will be used to address multiresource optimization. It will follow a probabilistic analysis and cover the main manufacturing flows along the circular cycle, which include energy, water and material. The optimization technique will minimize the consumption of primary resources and waste generation. The methodology will be based on the multiphysical modeling of the processes in collaboration with the client companies of the research group. Bio-based product manufacturing processes will be selected as test cases.
TECHNOLOGICAL INSTITUTE OF ARAGON
TOTAL BUDGET ITAINNOVA: 97,800 €.
TOTAL FINANCING ITANNOVA: 97,800 €.
The Juan de la Cierva – María Herrando action has been funded by the grant reference: IJC2020-043717-I- / MCIN/AEI / 10.13039/501100011033 and by the European Union “NextGenerationEU”/PRTR”, being IJC202020-043717- the reference that appears in the award resolution; MCIN the acronym of the Ministry of Science and Innovation; AEI the acronym of the State Research Agency; and 10.13039/501100011033 the DOI.
