The E2E DIGITAL TWIN project sets interconnected objectives to create an approach that not only improves the quality and efficiency of production but also modernizes worker training and introduces advanced work methods, significantly impacting both operational and strategic levels.
Project Summary:
The production of high-end natural cork stoppers relies heavily on visual inspections carried out by specialized workers to assess the quality of the stoppers at various stages of the production process. This evaluation impacts both the company’s profitability and customer satisfaction. If high-quality stoppers are mistakenly classified as defective, they are not sold. Conversely, if defective stoppers go undetected, they are sold, compromising customer trust. As a result, the value of the final product is highly dependent on the accuracy and efficiency of this process.
CorkSupply aims to modernize its methods to improve efficiency in quality assessment. This project seeks to disrupt the current inspection and classification process for natural cork stoppers by integrating digital and physical processes through:
- The creation of a Digital Twin for each high-end natural cork stopper produced, as well as for the inspection, quality control, and assembly processes they undergo.
- The use of Machine Learning to analyze and classify cork stoppers and stopper sets, reducing the overlap between adjacent quality classifications—higher (fewer imperfections) and lower (more imperfections).
- The use of Augmented Reality devices on the production line to combine workers’ tacit knowledge with data generated by mathematical models in the digital process. This ensures a meticulous, artisan-level quality control process—strategic for CorkSupply’s brand positioning—while improving efficiency and reliability.
- The development of technological strategies to characterize and understand the inspection processes performed by experienced workers and apply this knowledge to interactive multimedia training applications for new workers. This will streamline training processes and engage new generations.
Open positions
Doctoral student position #1
In the production of high end natural cork bottlestoppers, quality control and pricing is assured by several stages of visual inspection. This process involves craftsmanship as well as technology and automation, hence, presenting an interesting opportunity to evaluate the usability and effectiveness of applying augmented reality (AR) strategies to blend information derived from machine learning and computer vision processes to the workers expertise.
In this context, the objective of this project is to develop and evaluate AR applications to be used in the context of a natural cork bottlestopper production factory. The AR applications will be developed to fulfill distinct purposes, from assisting visual quality control inspections by overlaying digital annotations on top of real bottlestoppers, to tutorials and serious games for training new employees.
Work Plan
T1: Problem identification and characterization of the state-of-art in AR applications in industry.
T2: Preliminary research on the state-of-art of realtime object tracking using RGB and RGB-depth images.
T3: Development of prototype aplications to evaluate marker-based and markerless tracking technology in the task of real-time positional tracking of natural cork bottlestoppers.
T4: Development of AR-goggles applications that overlay pre-produced digital content on top of real bottlestoppers.
T5: Evaluate developed applications in lab and production settings.
T6: Results publishing.
More information and application:
https://euraxess.ec.europa.eu/jobs/323339
Application Deadline:
21 Mar 2025 – 23:59 (Europe/Lisbon)
Doctoral student position #2
In the production of high end natural cork bottlestoppers, quality control and pricing is assured by several stages of visual inspection. In this taks, experienced factory workers evaluate bottlestoppers and groups of bottlestoppers to detect specific defects and aesthetics indicators that will influence the price and overall acceptance of the product. This process involves craftsmanship as well as technology and automation, hence, presenting an interesting opportunity to evaluate the usability and effectiveness of applying augmented reality strategies to blend information derived from machine learning and computer vision processes to the workers expertise.
In this context, the objective of this project is to develop and evaluate augmented reality (AR) applications to be used in the context of a natural cork bottlestopper production factory. The AR applications will be developed to fulfill distinct purposes, from assisting visual quality control inspections by overlaying digital annotations on top of real bottlestoppers, to tutorials and serious games for training new employees.
Work Plan
T1: Problem identification and characterization of the state-of-art in AR applications in industry.
T2: Perform AR-goggle devices usability tests.
T4: Development of AR-goggles applications that overlay pre-produced digital content on top of real bottlestoppers.
T5: Evaluate developed applications in lab and production settings.
T6: Results publishing.
More information and application:
https://euraxess.ec.europa.eu/jobs/323321
Application Deadline:
21 Mar 2025 – 23:59 (Europe/Lisbon)
Postdoctoral Research #1
Current production of high quality natural cork bottlestoppers require several steps of visual and manual inspection implemented by specialized workers. Migrating this process to an automated strategy fundament by industrial digital twin concepts and machine learning (ML) and artificial intelligence (AI) algorithms will demand the generation of large annotated datasets.
In this project, the objective is to explore the possibility to use parametric and generative algorithms to create entirely digital, synthetic examples of natural cork bottlestoppers that can be used to train and test ML algorithms for performing automated quality control.
More information and application: https://euraxess.ec.europa.eu/jobs/320049
Application Deadline:
11 Mar 2025 – 23:59 (Europe/Lisbon)
Postdoctoral Research #2
Visual quality control (VQC) is a particularly relevant process in industries where the quality of the final product depends not only on functional aspects, but also on aesthetics. In such cases, human visual inspection is key for obtaining a high-quality product, which in turn, is conditioned by errors due to human factors such as fatigue and mood.
In the context of the digital transformation of a cork manufacturing industry, the objective of this grant is to develop visual discrimination experiments to evaluate the performance of real vs virtual natural cork bottle stoppers assessment (quality, classification, identification) using different virtual representations (3D models, 2D images) presented in different formats and mediums (smartphones, tablets, laptop, Tv screens, VR/AR goggles). The results of this experiments should guide the development of better modelling techniques to create the bottle stoppers’ digital twins, and also, a better understanding of the VQC process performed by the specialists.
More information and application:
https://euraxess.ec.europa.eu/jobs/320550
Application Deadline:
12 Mar 2025 – 23:59 (Europe/Lisbon)
Project details
Participating Institutions in the Project:
- Cork Supply Portugal, S.A.
- IPN – Instituto Pedro Nunes, Associação para a Inovação e Desenvolvimento em Ciência e Tecnologia
- Universidade de Coimbra
Funding/Managing Institution: Agência para o Desenvolvimento e Coesão
Funding Program: PT2030 – MPr-2023-07 – Aviso SIID – I&D Empresarial – Operações em Copromoção – Outros territórios
Execution period: 01/01/2024 to 02/04/2026
