1,720,969 research outputs found
Collaborative robots in quality control and management: applications and challenges
Full text linkL'abstract è presente nell'allegato / the abstract is in the attachmen
Preliminary comparison between manual assembly and intelligent human-robot collaborative assemblies in terms of quality and assembly time
Full text linkIn recent years, the use of Human-Robot Collaboration (HRC) in manufacturing systems has grown significantly, within the framework of Industry 4.0 and emerging Industry 5.0. Collaborative robots, thanks to their ability to reduce physical and mental stress of operators, enable increased productivity and quality performance. This paper analyses assembly time and quality trends as a function of assembly complexity in intelligent collaborative assembly and makes a holistic comparison between a manual assembly and two different collaborative assemblies, focusing on assembly times and in-process errors. The assembly of products with different levels of complexity is used as a case study
Integrative inspection methodology for enhanced PCB remanufacturing using artificial intelligence
No full textElectronic waste (e-waste) represents one of the world's most significant environmental challenges, with over 50 million tons generated annually. A key component is the management of Printed Circuit Boards (PCBs), which are integral components of electronic devices and have an operational lifespan of 15 years. However, on average, electrical equipment is discarded after 5 years due to individual defects, prompting the EU to enforce regulations supporting the right to repair. Although industrial remanufacturing of PCBs could be a viable solution, it is not currently feasible due to the complex inspection process required. This paper presents a novel inspection process approach based on data fusion of thermography, current measurement and optical inspection using artificial intelligence. The result is intelligent diagnostics in less time and with lower investment costs. In addition to the concept, initial investigations with real industrial applications in the field of automation are presented
Challenges and opportunities of collaborative robots for quality control in manufacturing: evidences from research and industry
Full text linkPurpose - In the context of Industry 4.0, collaborative robots - which might be equipped with different types of sensors - have been gaining ground, used to cooperate with humans in quality control of finished or semi-finished products. Compared to the various applications of collaborative robotics in manufacturing (e.g., material handling, assembly, pick and place, and positioning), widely studied and adopted in industry, quality control and testing have not yet reached their full potential. This paper aims to study the state-of-the-art collaborative robotics used for quality control purposes in both academia and industry.
Design/methodology/approach – This paper analyses in a structured way the scientific literature and some prominent real industrial case studies regarding the state-of-the-art of quality control using collaborative robotic systems in manufacturing.
Findings - The analysis enables the identification and definition of the main challenges and opportunities that the manufacturing sector is facing for the large-scale use of the new quality control paradigm. Results show that collaborative robotics in quality still plays a marginal role and is mainly adopted for in-process visual inspections to increase system efficiency. Some barriers still hamper the full adoption of this paradigm, but there is plenty of opportunity for research and economic growth.
Originality/value - The innovative aspect of this research is the combined analysis of scientific articles and real-life case studies that provide a comprehensive overview of the research and actual use in industry of this emerging paradigm of quality control
Toward a concept of digital twin for monitoring assembly and disassembly processes
Full text linkQuality is one of the key factors in the customer’s selection process between competing products. Producing high-quality, defect-free products that meet consumer expectations is crucial for manufacturing companies to gain a competitive advantage. Accordingly, developing appropriate defect generation models is essential in modern manufacturing companies to predict defects and plan efficient quality control and production. On the other hand, with its ability to support new business models and decision support systems, digital twin technology is one of the new technologies emerging to support digital transformation. Faster optimization algorithms, more powerful computers, and a massive increase in available data are just some of the features of digital twins that can be used to advance simulation toward real-time quality control and optimization of products and production systems. This paper aims to model the generation of defects of product variants in assembly and disassembly processes and evaluate their integration within a digital twin system to prevent the occurrence of defects and ensure product quality. The proposed strategy is expected to improve the optimization, monitoring, and diagnostic capabilities of complex product variants’ assembly and disassembly systems, realizing an upgrade from a single physical implementation to a combination of physical and digital
Automatic component recognition and defect detection in electronic board recycling process
No full textThe growing problem of Waste Electrical and Electronic Equipment (WEEE), or e-waste, posessignificant environmental and resource challenges that require innovative management strategies.The main objective of this research is to develop an automated system that can detect defects inelectronic components, enabling the reuse of electronic boards and reducing their environmentalfootprint. A comprehensive methodology using Convolutional Neural Networks (CNNs) and MachineLearning (ML) is proposed, targeting to inspect different customised boards with different compo-nent layouts. The approach exploits the capabilities of advanced pattern recognition and predictiveanalysis to identify faults in electronic components. To preliminarily validate and demonstrate theeffectiveness of this methodology, a simple case study was considered. Extensive testing on this casestudy confirmed the potential of the method, achieving a 95% confidence level of defect detection.The proposed methodology aims to extend the life of electronic devices, improve maintenancestrategies and promote sustainable use. This strategic application addresses the current challenges ofmanaging e-waste and paves the way for future advances in managing it
Automatic object detection for disassembly and recycling of electronic board components
Full text linkThis paper presents the development of a deep learning-based object recognition system designed to automate and speeding up the disassembly process of electrical and electronic components. The main goal is to address the mounting global issue of Waste Electrical and Electronic Equipment (WEEE) management. The increasing availability of technology and the expansion of consumer markets have led to a significant surge in the generation of WEEE, necessitating the urgent development of sustainable and automated strategies for its disposal and resource recovery. Traditional manual disposal methods and the uncontrolled accumulation of WEEE can pose serious threats to the environment, human health and natural resources. A comprehensive approach, involving advanced recycling technologies, life-cycle management and policy reforms is required to handle this escalating waste stream.
The complexity of WEEE management is heightened by the diversity of product design and composition, making efficient material selection processes often labour-intensive and expensive. This work focuses on the automated detection and sorting of WEEE products. The proposed system enables rapid identification of products and components, in order to facilitate the subsequent disassembly and reuse of components that are still considered functional, safe, and of good quality. This concept is illustrated through a case study where recognition of six different electronic boards is performed
Improved quality control and sustainability in food production by machine learning
Full text linkIn recent years, the food industry has faced a number of complex challenges related to both quality control and sustainability. Ensuring consumer safety and satisfaction remains a cornerstone of the food industry, supported by stringent standards that address the risks of contamination and spoilage. However, variability in raw materials, processing techniques and storage conditions are just some of the factors that affect quality in the food industry. To manage this high variability, it is essential to analyse the production process and factors that most influence food quality, aiming to predict and minimise food waste, thereby ensuring a sustainable process. This convergence of quality control and sustainability goals provides fertile ground for machine learning applications. By improving defect detection, process optimisation, resource allocation and predictive maintenance, these models help to improve product quality and reduce environmental impact. This article aims to explore the various applications of machine learning models in the food industry, where the variability of raw materials and the difficulty of controlling production and environmental factors challenge the use of traditional methods. The quality control and sustainability of an industrial corn cakes production process is used as a case study
Subjective vs objective assembly complexity assessment: a comparative study in a Human-Robot Collaboration framework
Full text linkThe impact of manufacturing complexity on company performance can be significant, affecting productivity, efficiency, affordability, and quality if not managed correctly. Assessing and managing manufacturing complexity is a multifaceted task that involves both objective and subjective features, such as product complexity, assembly sequence, operator factors, and operation/management strategies. This study proposes a structured methodology to assess the perceived complexity of human-robot collaboration assembly processes. The methodology is based on 16 assembly complexity criteria and a multi-expert decision-making method for evaluation. Operators assign importance scores and agreement levels to each criterion using a five-level ordinal scale, and the linguistic data is processed using the Multi-Expert Multi-Criteria Decision Making (ME-MCDM) method [Yager(1993)]. This approach combines linguistic information provided for non-equally important criteria using maximum, minimum, and negation operators to obtain an overall synthetic linguistic value of perceived complexity using fuzzy logic. The proposed approach provides an assessment of perceived complexity at both individual and overall levels, aggregating all individual complexity assessments by the operator Ordered Weighted Average (OWA) [Yager(1993); Filev(1994)]. The proposed approach for assessing perceived complexity of assembly is compared with a purely objective assessment method, firstly proposed by Sinha et al. [Sinha(2012)]. This model was validated in various studies, and its effectiveness in quantifying the complexity of industrial products was demonstrated [Verna(2022)]. It is based on the molecular orbital theory and is applied to the engineering domain to analyse the complexity of cyber-physical systems. The model represents a cyber-physical system as several connected components where each component can be thought of as an atom, and the interfaces between them as inter-atomic interactions or chemical bonds. The complexity of the assembly is defined as the combination of three complexity components: handling complexity (C1), connections complexity (C2), and topological complexity (C3), as follows C=C1+C2∙C3. This objective model, based on structural characteristics of the assembly process, was used as a reference model for the subjective complexity model.
The comparison between subjective and objective assessment of complexity was performed in a real-world production environment, using a human-robot collaboration process for manufacturing custom electronic boards with different levels of complexity. The results showed a significant correlation between individual perceived complexity and objective complexity, indicating that the proposed perceived complexity model can be linked to the objective model. As structural complexity increases, higher levels of individual perceived complexity become more likely, but the variability in perceived complexity varies with structural complexity. These findings suggest that individual operator ability and cognitive factors, such as training, knowledge, and cultural and organisational factors, play a role in perceived complexity and require further investigation. The study also suggests that using perceived complexity to assess assembly complexity is suitable for low- and medium-complexity products, but not for high-complexity products, where objective complexity models may be more appropriate, since after a certain point operators do not distinguish between different levels of complexity.
The proposed methodology and data analysis approach offer a new perspective on assessing perceived complexity, relying solely on synthesis operators and statistical tools suitable for categorical data. Engineers can use the study's results to minimise perceived complexity and ensure alignment between perceived and objective complexity
Real‐Time Monitoring of Human and Process Performance Parameters in Collaborative Assembly Systems using Multivariate Control Charts
Full text linkWith the rise in customized product demands, the production of small batches with a wide variety of products is becoming more common. A high degree of fexibility is required from operators to manage changes in volumes and products, which has led to the use of Human-Robot Collaboration (HRC) systems for custom manufacturing. However, this variety introduces complexity that afects production time, cost, and quality. To address this issue, multivariate control charts are used as diagnostic tools to evaluate the stability of several parameters related to both product/process and human well-being in HRC systems. These key parameters monitored include assembly time, quality control time, total defects, and operator stress, providing a more holistic view of system performance. Real-time monitoring of process performance along with humanrelated factors, which is rarely considered in statistical process control, provides comprehensive stability control over all customized product variants produced in the HRC system. The proposed approach includes defning the parameters to be monitored, constructing control charts, collecting data after product variant assembly, and verifying that the set of parameters is under control via control charts. This increases the system's responsiveness to both process inefciencies and human well-being. The procedure can be automated by embedding control chart routines in the software of the HRC system or its digital twin, without adding additional tasks to the operator's workload. Its practicality and efectiveness are evidenced in custom electronic board assembly, highlighting its role in optimizing HRC system performance
- …
