Conference Agenda

New technologies related to Digital Transformation (DT) and the Industry 4.0 (I4.0) modify the way business and productive processes are carried out, generating complex changes for industry and engineering, establishing new tasks and human roles, and interacting with the characteristics of Transdisciplinary. Digital engineering managers play an integrative role by relating and using the organisation's digital technological knowledge to generate better business results. The characterization of managers' competences to guide and stimulate value creation in industrial sectors is still not sufficiently investigated and emerges as a critical element for industrial development in the digital age. This research fulfils this gap and aims to rank four types of necessary competences for engineering managers facing the DT/I4.0 in the automotive sector. The methodological approach adopted is quantitative, based on the judgement of engineering managers from the Brazilian automotive sector, which is globally representative in terms of productivity. An Analytic Hierarchy Process (AHP) is applied in the data treatment. Results are based on a sample of 35 interviews from six automotive companies with different levels of complexity in production operations and formal programs for DT/I4.0 implementation. Findings indicate the relative priority for the digital technical, managerial, social, and motivational competences, presenting insights with implications to guide the development of the digital engineering managers.

The financial market stress of 2008, triggered by the collapse of Lehman Brothers, underscored the critical need for efficient collateral management in financial transactions. This research focuses on developing and validating a blockchain platform designed to automate the collateral value adjustment in securities lending transactions. The platform proposes innovative transaction flows and algorithms to manage collateral automatically, significantly making risk management more effective and reducing administrative costs. The study’s methodology includes 1) designing a blockchain platform for automated collateral management, 2) developing and testing collateral value adjustment algorithms, and 3) conducting case studies with real stock and bond price data to evaluate the platform’s effectiveness. A notable outcome of this research is the reduced credit risk and liquidity compared to traditional procedures, based on the benefit of collateral diversification enabled by the automated algorithms employing multiple tokens as collaterals. This research contributes to the field of transdisciplinary engineering by integrating financial and technological disciplines, particularly in leveraging blockchain technology for financial applications. Furthermore, the study reflects on broader societal impacts, such as improving the efficiency and security of financial transactions, potentially stabilizing market volatility, and offering insights for policy-making in sound and robust settlement systems.

The ongoing digitalisation of manufacturing shop floors generates an abundance of valuable data, providing novel opportunities for real-time monitoring of performance data. Digital dashboards, equipped with technology-enabled access to central details and tailored to practical shop floor settings, possess functionalities to make a significant impact in this context. The effectiveness of digital dashboards depends on the extent to which the provided functionalities meet the requirements of the shop floor. In this intervention-based research, we utilize the theory of constraints to identify and address constraints in data and information accessibility, measurement methods, and the physical flow of materials. The study subscribes to transdisciplinary research, drawing on operations management and technology management theories. We propose a gradual approach to identify and address constraints, serving as guidelines for designing digital monitoring systems aimed at achieving throughput-oriented shop floors.

Predicting product quality is a crucial element in smart manufacturing. Then, the development of robust learning models is essential based on historical data relevant to their respective products and production processes. The challenge of quality prediction models lies in the limited availability of production batches and their associated historical data in highly customized products like large power transformers, precision machine tools, and other industrial equipment. Despite the small dataset, ensuring the quality of the final product remains paramount. This study introduces an innovative transfer learning approach integrating adaptive machine learning and non-linear regression. The goal is to accurately predict the quality of highly customized products using datasets limited by the constraints of the primary suppliers with smaller production scales. The research employs a case study and dataset from the production of large power transformers. The input data for training and testing the predictive model include key power transformer parameters, such as core loss values and power loss. The approach utilizes transfer learning that transfers knowledge gained from one task to a similar task. The proposed method enhances the model's performance and generalization capabilities. Subsequently, the model is fine-tuned rapidly without compromising accuracy. This paper contributes to a comparative analysis with previous research, demonstrating the effectiveness and superiority of the proposed method. Manufacturers can leverage this approach to predict the quality of complex, small-batch, and highly customized industrial products. Ultimately, this method aids in improving production quality and reducing costs.