ICAC2023 – The Chinese Automation and Computing Society in the UK

Title:

Smart Manufacturing Review and Perspectives

Abstract:

Smart manufacturing also known as intelligent manufacturing has been recognized as major driving force to transform manufacturing industry. The first international collaboration on Intelligent Manufacturing Systems (IMS) was proposed by Japan with participation of US, Canada, Australia, and European partners in early 90s. In the following years, artificial intelligence (AI) applications in manufacturing were limited to certain focused areas. Since Germany introduced Industry 4.0, other major manufacturing nations such as US, China and Japan proposed programs to promote advanced and smart manufacturing. The recent development of ChatGPT demonstrated the potential of AI technologies. The society started to realize that AI will have profound impact not only on manufacturing industry, but also almost all other sectors of the society. This speech will provide a brief historical review of IMS, the current technological development of intelligent product design and manufacturing, and future perspectives of smart manufacturing in digitalization technologies and industrial metaverse.

Biography:

Peihua Gu is currently President of International Institute for Innovative Design and Intelligent Manufacturing of Tianjin University in Zhejiang China, a Professor of Mechanical Engineering and Executive Director of Center for Emerging Engineering Education Research and Application, Tianjin University. He was Provost and Vice-President of Shantou University from 2005-2018. Prior to returning to China, he was Professor and Head of the Department of Mechanical and Manufacturing Engineering, the University of Calgary, Canada.

He was an elected Fellow of Canadian Academy of Engineering (2004) and Fellow of CIRP (2004). He was twice awarded Industrial Research and Design Chair by the Natural Science and Engineering Research Council of Canada (NSERC) and was Changjiang Scholar Chair Professor by Ministry of Education of China. His main research contribution includes the establishment of Adaptable Design Method, invention of Multi-material and Multi-source Deposition Method and various techniques in intelligent manufacturing. He is an author and co-author of over 300 technical publications. He received several research and education awards and delivered a number of keynote and invited lectures.

Professor Gu served on NSERC of Canada and NSF of China and the Editorial Boards for over 10 Journals as well as other research and education organizations. He received his PhD from McMaster University, M.Eng. and B.Eng (Dip) from Tianjin University.

Title:

Explainable AI: Intelligent Inference and Evidence-based Decision Making – the Evidential Reasoning Perspective

Abstract:

This presentation is dedicated to reporting the findings from a number of recent and current research projects in the areas of explainable AI (XAI) and interpretable machine learning (IML) for developing XAI decision support systems (XAI-DSSs) to assist evidence-based decision making in professional services, in particular claim handling in insurance and loan underwriting in finance. Our understandings of XAI and IML are first discussed in relation to probabilistic inference by briefly describing Bayes’ rule, Dempster’s rule and the evidential reasoning (ER) rule and their relations, and also to intelligent modeling based on both historical data and human knowledge by describing the belief-rule-base (BRB) methodology. A couple of examples in engineering systems (e.g. smart car manufacturing, system risk analysis and fault diagnosis) and the key features of some real world BRB-oriented and ER-powered XAI-DSSs in professional services are then presented to elaborate the descriptions. These XAI-DSSs can both accommodate human know ledge and learn from historical data via IML. They consist of factual and heuristic rules of a probabilistic nature, each of which is explainable to experienced decision makers (e.g. underwriters, claim handlers and lawyers). These probabilistic rules can be combined to infer decisions that are also probabilistic. The systems can provide explanations to rejected or accepted claims and loan applications by following a chain of probabilistic rules leading to the decisions. The working logic of the systems can be easily narrated and understood by non-technical people. It can be utilized to provide prompt advice to claim handlers and underwriters to make fast and robust decisions. The business case studies and implementation processes show that the XAI-DSSs provide explainability and interpretability that help build trust among stakeholders, which is vital for acceptance of any AI decision support systems in professional services.

Biography:

Dr Jian-Bo Yang is Chair Professor of Decision and Systems Sciences at Alliance Manchester Business School, The University of Manchester, UK. Over the last three decades, He has conducted theoretical and methodological research in many areas, including the evidential reasoning (ER) theory; multiple criteria decision analysis under uncertainties; probabilistic inference and decision analysis using both data and judgments; multiple objective optimisation; intelligent system modelling; explainable artificial intelligence (XAI) and interpretable machine learning (IML); hybrid decision methodologies and technologies combining concepts and techniques from decision science, systems science, operational research and AI. His current applied research covers a range of applications driven by historical data, enabled by human knowledge and powered by ER, including modelling and decision support for professional services such as finance, insurance and healthcare; diagnosis and prognosis, design and operation decision analysis in healthcare, engineering and social systems; pattern identification and analysis of consumer behaviours; analysis of public sentiments and system risks (financial or non-financial); new product development; aggregated production management; system maintenance management; risk and security modelling and analysis; performance analysis and improvement of products, processes and organizations. He is principal investigator or co-investigator for over 70 research projects at a total value of over £21m, funded by many organisations, including UK Engineering and Physical Science Research Council (EPSRC), UK Economic and Social Research Council (ESRC), Innovate UK, UK Department of Environment, Food and Rural Affairs (DEFRA), European Commission (EC), Natural Science Foundation of China (NSFC), Hong Kong Research Grant Council (HKRGC), US Government and industry. He has published 4 books, over 270 journal papers and book chapters, and a similar number of conference papers, with extensive citations in Web of Science and Google Scholar, and developed several software packages in optimisation and decision support with wide applications worldwide.