The correct maintenance of industrial machines is critical to achieve the zero defects manufacturing paradigm promised by Industry4.0. In this context, AR (Augmented Reality) is a key enabler to overcome this challenge. The ability to see information overlays on top of industrial machines has the potential to completely transform the areas of training and maintenance contributing to the zero-defect objective. The main goal of this sub-project is to bring an AR service into the ZDMP platform. A new zComponent will be developed and integrated that will facilitate ZDMP platform end users to use AR applications to visualize ZDMP platform data.

Development of Robot Kinematics Component for bringing robust Internet of Robotics Things and Cloud Robotics applications to Zero-Defects Manufacturing Platform. ZDMP will be able to offer a way for robot basic mathematical modelling and motion planning validation other components as for instance Machine Learning algorithms. This component will allow to model and validate robot’s trajectories preventing faults, increasing robustness, and optimizing the production process in zero-defects applications.

In the last decades computer vision has developed many standardized features to describe objects represented in video streams; modern industrial defect recognition systems based on camera sources can thus easily take advantage of machine learning techniques for classification. On the contrary no such a standardized feature-based description of structural 3D objects is currently available. The goal of the proposed project is to support developing defect classifiers based on both 3D design-time models (CAD) and 3D object data, computing and exposing a complete set of well-defined and manufacturing-oriented 3D object descriptors.

MX3D proposes a solution to address consistency problems in robotic AM. Universal Calibration Tool for Robotic Additive Manufacturing (UCRAM) will be a tool that monitors the layer build-up, either to measure and calibrate the material build-up beforehand during a calibration procedure, or monitoring it layer-by-layer as the print progresses. The tool will create a digital twin of the printed part, by scanning each layer using a 2D-laser scanner, thus building up a 3D-image that can be used for review or to overlay with other available data sets. A simple, fast, and low-cost tool for defect detection and quality control in any type of robotic AM to achieve zero-defect additive manufacturing.

Quality control is essential for a manufacturing company to develop a sustainable business, not only from an environmental point of view but also socially and economically. Thus, it is essential to facilitate integrating a platform such as ZDMP in industrial facilities where existing software can do that. This project aims to integrate RAILES platform responsible for making some functionalities on the shop floor with ZDMP. RAILES can be used to extract data and send it to ZDMP. The consortium intends to explore the integration of ZDMP with software platforms already running at the factory, reducing the risks of integrating these solutions, which is usually seen as a considerable barrier.

Based on our ongoing work in multivariate data analytics for zero defect manufacturing and current work on cognition-driven systems for improving process quality, we propose a novel approach for ZDM driven by a proactive detection of complex variations in processes, their understanding (root-causes) and impact analysis, with the goal to spot unusual behaviour of a process and validate it, before it starts producing defect products. The main advantage is that the approach is focused on monitoring and understanding process behaviour (as a whole) and not “isolated” anomalies. Outcome is an innovative ZDMP-enabled solution, cognition-driven ZDM (CogniZDM), based on existing ZDM components and integrated in the ZDMP Platform.

The present proposal falls into the scope of validation sub-projects, intending to demonstrate the usefulness of 14 ZDMP components. ZDMP open call comprises a unique opportunity for Pressious to establish the principles of Industry 4.0 under the circular economy paradigm, in terms of proactive resource management, zero-defect manufacturing, and environmental footprint minimization, by exploiting already available datasets across the production chain. The output of this project will be a thorough examination, testing and validation of these components in the industrial domain of offset printing and the potential adoption of ZDMP platform after the project lifetime.

In our study on 100 dental implants (80 producers) we found 30% of the products contaminated with particles, hazardous to human health and not compliant to requirements in clean production. Our solution on ZDMP provides a standardised examination method based on electron microscopy and fluorescence imaging. We provide a UI and toolsets for image analysis on the local machine, referencing to our verified data catalogue of dental implant contaminants. Utilising AI and ML analysis the user will receive a cleanliness evaluation and a trusted quality seal for each product under examination at his premise. The method can be translated to other metallic or ceramic materials.

Factories in the transformation process to I4.0 require to involve employees in production processes who thus are a crucial factor for reaching zero-defect. In this sub-project, a new zComponent will enable the collection of unstructured data in manufacturing. The proposed zComponent provides an API allowing requests from external systems as well as the manual creation of checklists. Thus dynamically created user input forms are provided in a mobile app. Results are converted into a semantic structure and returned to the requesting system and platform by means of smart processing algorithms.

Zero Defect Manufacturing (ZDM) is intrinsically a multi-disciplinary approach requiring parallel progress in domains such as data analytics and knowledge management. Data that capture the knowledge and experience of people working in manufacturing, addressing and analysing quality issues can be combined with typical machine analytics data. Industrial Social Network (ISN) solutions can be very efficient for knowledge sharing in manufacturing. The objective of ISN4ZDM is the integration of RAPpID ISN solution to ZDMP. RAPpID complements existing ZDMP functionalities with new capabilities that are geared towards management of collaborative ZDM knowledge combined with IIoT data.