Periodic Reporting for period 1 - Platform-ZERO (Customizable AI-based in-line process monitoring platform for achieving zero-defect manufacturing in the PV industry)
Periodo di rendicontazione: 2023-01-01 al 2024-06-30
The Platform-ZERO project addresses a critical issue in the photovoltaic (PV) industry, particularly in third-generation PV technologies, which are essential for advancing green energy transitions and combating climate change. These technologies, while highly efficient and customizable, have complex production processes that make them prone to defects and small deviations during manufacturing that result in low-performance devices, increased costs and limited sustainability.
To address these critical challenges the project’s objective is to develop a modular, self-learning platform for real-time production monitoring powered by artificial intelligence (AI) to reduce production defects and deviations, lower production costs, and improve PV device performance. The developed platform will detect and correct pre-critical failures, preventing waste and ensuring more efficient manufacturing processes. Proposed modular design of the platform allows its application in various industries beyond PV, including semiconductors and chemicals, expanding its potential impact. By improving the quality and efficiency of PV devices, Platform-ZERO will facilitate their integration into applications such as building-integrated photovoltaics, vehicle-integrated photovoltaics, and agrivoltaics. These improvements will make solar energy more accessible and affordable, helping to drive Europe’s energy transition. Additionally, the project enhances Europe’s energy security by enabling the production of clean, renewable energy at a lower cost, reducing reliance on non-renewable resources. Application of AI-based solutions align with the broader goals of Industry 4.0 crucial for the continent's green and digital transitions.
The project also aims to propose new standards for production process monitoring, addressing current limitations and fostering innovation. This will enhance Europe's industrial competitiveness, especially as the global market increasingly values digitalization and sustainability.
In conclusion, Platform-ZERO tackles a critical challenge in third-generation PV production through AI and data management. Its innovations extend far beyond PV, supporting climate neutrality, industrial digitalization, and job creation across Europe. This holistic approach not only addresses production inefficiencies but also generates long-term social and economic benefits, reinforcing Europe's leadership in the green energy transition and the digital age.
https://www.platform-zero-project.eu/(si apre in una nuova finestra)
To address these critical challenges the project’s objective is to develop a modular, self-learning platform for real-time production monitoring powered by artificial intelligence (AI) to reduce production defects and deviations, lower production costs, and improve PV device performance. The developed platform will detect and correct pre-critical failures, preventing waste and ensuring more efficient manufacturing processes. Proposed modular design of the platform allows its application in various industries beyond PV, including semiconductors and chemicals, expanding its potential impact. By improving the quality and efficiency of PV devices, Platform-ZERO will facilitate their integration into applications such as building-integrated photovoltaics, vehicle-integrated photovoltaics, and agrivoltaics. These improvements will make solar energy more accessible and affordable, helping to drive Europe’s energy transition. Additionally, the project enhances Europe’s energy security by enabling the production of clean, renewable energy at a lower cost, reducing reliance on non-renewable resources. Application of AI-based solutions align with the broader goals of Industry 4.0 crucial for the continent's green and digital transitions.
The project also aims to propose new standards for production process monitoring, addressing current limitations and fostering innovation. This will enhance Europe's industrial competitiveness, especially as the global market increasingly values digitalization and sustainability.
In conclusion, Platform-ZERO tackles a critical challenge in third-generation PV production through AI and data management. Its innovations extend far beyond PV, supporting climate neutrality, industrial digitalization, and job creation across Europe. This holistic approach not only addresses production inefficiencies but also generates long-term social and economic benefits, reinforcing Europe's leadership in the green energy transition and the digital age.
https://www.platform-zero-project.eu/(si apre in una nuova finestra)
During the first period of the Platform-ZERO project (months 1-18), significant progress was made toward the project’s objectives. The partially achieved results can be categorized into four key areas essential for implementing the proposed monitoring platform: hardware development for industrial process monitoring, automation systems for operations, AI algorithm design for big data interpretation, and the development of a robust infrastructure for big data management.
Hardware development: Specialized sensors are being designed to transfer laboratory characterization capabilities to an industrial environment. This includes sensors based on spectroscopic, imaging, and optoelectronic techniques. Strategies for sensor fusion and simultaneous operation, along with automatic calibration algorithms, are also being developed.
Automation of operations: Modular and customizable hardware and software systems are being developed to integrate and automate inspection sensors in roll-to-roll (RtR) and sheet-to-sheet (StS) production processes. These systems enable real-time inspection and quality control, accelerating the digitalization of industrial activities, improving production efficiency, and enhancing product quality.
Artificial intelligence algorithms: A key challenge of the project is managing the large volumes of data generated by sensors. A first version of an AI architecture based on machine learning has been developed, automating the learning of industrial processes and enabling efficient monitoring and quality control in a semi-supervised environment. This rapid data processing will allow for early detection of deviations and optimization of processes.
Big data management: The project is creating a big data management system for holistic and customized inspection of industrial processes. This infrastructure will enable better data management and the creation of data standards.
These advances lay the groundwork for the implementation and validation of the holistic platform for the production process monitoring in PV and related industries.
Hardware development: Specialized sensors are being designed to transfer laboratory characterization capabilities to an industrial environment. This includes sensors based on spectroscopic, imaging, and optoelectronic techniques. Strategies for sensor fusion and simultaneous operation, along with automatic calibration algorithms, are also being developed.
Automation of operations: Modular and customizable hardware and software systems are being developed to integrate and automate inspection sensors in roll-to-roll (RtR) and sheet-to-sheet (StS) production processes. These systems enable real-time inspection and quality control, accelerating the digitalization of industrial activities, improving production efficiency, and enhancing product quality.
Artificial intelligence algorithms: A key challenge of the project is managing the large volumes of data generated by sensors. A first version of an AI architecture based on machine learning has been developed, automating the learning of industrial processes and enabling efficient monitoring and quality control in a semi-supervised environment. This rapid data processing will allow for early detection of deviations and optimization of processes.
Big data management: The project is creating a big data management system for holistic and customized inspection of industrial processes. This infrastructure will enable better data management and the creation of data standards.
These advances lay the groundwork for the implementation and validation of the holistic platform for the production process monitoring in PV and related industries.
The current state of industrial process monitoring in the semiconductor industry in general, and in PV industry in particularly, is limited by the use of in-line inspection techniques such as computer vision, photoluminescence, X-ray fluorescence, reflectance, transmittance, and ellipsometry. While these techniques improve inspection capabilities during production, they fail to provide a fully comprehensive evaluation of complex optoelectronic devices such as thin film PV modules. For a more precise and detailed inspection of produced materials and devices, advanced methods such as Raman spectroscopy, time-resolved photoluminescence, external quantum efficiency measurements, and spatially resolved imaging are required. However, these techniques are often only available in advanced laboratories with slow response times, which are insufficient for meeting the rapid feedback requirements of industrial production.
Another key challenge is that measurements must be performed in a single location to obtain a complete picture of materials and devices properties. Currently the measurement information is fragmentated between various production steps and the reliance on off-line analysis of selected products limit the ability to fully optimize and evaluate the process, delaying defect detection and preventing timely adjustments of manufacturing parameters.
In many cases the current control of the fabrication process is focused on production parameters, like temperature, pressure, gas flow, and deposition rate, rather than on the final product quality. This hampers real-time detection and correction of defects, as deviations in process parameters do not always correlate with defects in the finished product. Additionally, managing large volumes of data generated during production is a significant challenge. Pre-established calibrations, which require specific optimization for each industrial process, involve high implementation time and cost, slowing the response to changes in production environments.
To overcome all these challenges development of a holistic system that combines real-time monitoring of process parameters and product quality at different production steps is crucial. This would enable early defect detection and more efficient production optimization, especially for third-generation PV technologies, which are highly sensitive to minor parameter variations. Real-time data processing and calibration adjustment are essential for competitive, efficient production that meets modern industry demands.
The Platform-ZERO project advances existing technologies by providing a comprehensive, real-time monitoring and control platform that integrates AI and advanced sensor technology, enabling dynamic responses to production anomalies and setting a new standard for manufacturing oversight.
Another key challenge is that measurements must be performed in a single location to obtain a complete picture of materials and devices properties. Currently the measurement information is fragmentated between various production steps and the reliance on off-line analysis of selected products limit the ability to fully optimize and evaluate the process, delaying defect detection and preventing timely adjustments of manufacturing parameters.
In many cases the current control of the fabrication process is focused on production parameters, like temperature, pressure, gas flow, and deposition rate, rather than on the final product quality. This hampers real-time detection and correction of defects, as deviations in process parameters do not always correlate with defects in the finished product. Additionally, managing large volumes of data generated during production is a significant challenge. Pre-established calibrations, which require specific optimization for each industrial process, involve high implementation time and cost, slowing the response to changes in production environments.
To overcome all these challenges development of a holistic system that combines real-time monitoring of process parameters and product quality at different production steps is crucial. This would enable early defect detection and more efficient production optimization, especially for third-generation PV technologies, which are highly sensitive to minor parameter variations. Real-time data processing and calibration adjustment are essential for competitive, efficient production that meets modern industry demands.
The Platform-ZERO project advances existing technologies by providing a comprehensive, real-time monitoring and control platform that integrates AI and advanced sensor technology, enabling dynamic responses to production anomalies and setting a new standard for manufacturing oversight.