Computer Science: Fully Funded Swansea and Grenoble Joint PhD Scholarship: Hybrid AI for Supply Chain Optimisation in Industry 4.0
Funding providers: Swansea University and Université Grenoble Alpes
Subject areas: Computer Science
Project start date:
- 1 October 2022 (Enrolment open from mid-September)
Project supervisors:
- Professor Arnold Beckmann - Swansea University
- Dr Abdourahim Sylia - Université Grenoble Alpes, DCM
Aligned programme of study: PhD in Computer Science
Mode of study: Full-time
Project description:
Digital transformation is happening everywhere in our society. In the context of Industry 4.0 (I4.0), interconnected digitalised assets like Industrial Cyber-Physical System constantly interact with each other, and with humans with appropriate digital interfaces. Predictive and prescriptive data analytics utilising various methods from AI provide key technologies to make step changes in this context. It has been recognised that the human, and society in general, should be the focus of any such transformation [Industry 5.0 - Towards a sustainable, human- centric and resilient European industry. EU Policy Document, 2021].
Therefore, any technology developed in this context should be able to capture human knowledge, present derived knowledge in a human readable way, and, in general, be able to explain decisions to the human involved. Symbolic AI, in form of Expert Systems and more general Knowledge Based Systems, have shown successes to deal with knowledge in the early days of AI, but struggled with unstructured data. Statistical AI, in form of neural networks, machine learning and deep inference systems, have been very successful in recent years with unstructured data, but derived models are difficult to understand and reason about. Recently, Hybrid AI has been proposed, which combines sound symbolic reasoning with efficient machine learning to overcome those challenges.
This project considers supply chain optimisation and resilience challenges using Hybrid AI, to provide a framework for prediction and optimisation of supply chains. Potential use cases include:
- Supply chain risk management (at the supply chain network level): Covid pandemics have proven that all industrial supply chains are subject to risks. Predictive analytics is of particular importance to mitigate risks.
- Production planning under uncertainties (at the node level in a supply chain), where the focus is on a single node of the supply chain, for example operational risks at manufacturing plant level (machine break down, human errors, etc.) by applying predictive analytics for production planning.
The project is a joint project between Swansea University (SU) and Université Grenoble Alpes (UGA), with primary supervisors Professor Arnold Beckmann (SU) and Dr Abdourahim Sylia (UGA), and Secondary Supervisors Dr Cinzia Giannetti (SU) and Professor Gülgün Alpan (UGA). The project will be conducted 18 months at Swansea University, Computational Foundry, and 18 months at UGA, the I-MEP2 doctoral school which specialises in Industrial engineering.
Interviews will be held from 30 June.