Resilience and Robustness of Dynamic Manufacturing Supply Networks

Efficient and effective manufacturing supply networks (MSN) are essential to the functioning of the global economy. In line with the EPSRC call, this proposal is premised on the strong belief that appropriate mathematical theory and methods can provide fundamentally new understanding on the behaviour of MSNs and provide an effective investigative toolset for MSN analysis, design and management. In particular we argue that the power of network science can be harnessed to underpin new thinking in MSNs for resilience and robustness.

The work will be strongly embedded in real MSNs in three domains - producer-driven inbound MSNs and outbound distribution channels for industrial companies; global MSNs for critical products used in high-valued manufacturing (e.g. titanium or composite pre-preg materials); and evolving MSNs for emerging UK industries such as renewable energy. The project will develop and apply existing and new mathematics specifically in the theory of complex adaptive networks, drawing on techniques from game theory, dynamical systems and Bayesian informatics. It will also learn from related modelling approaches in ecology, metabolism modelling and utility grids.

This grant will represent the first attempt to develop an integrated mathematical modelling suite to support effective decision making in MSNs in the context of risk and uncertainty. The work will build on disparate recent developments in network science and complex adaptive dynamical systems, Bayesian statistics and operational research to develop new models and measures to better understand and analyse MSN behaviour and performance. Multiple perspectives and a multi-level view of risks and vulnerabilities in MSNs will be taken, including physical, financial, informational, relational, and governance perspectives at the strategic MSN design and policy levels, and risk mitigating strategies at both strategic and operational levels to support MSN management.

This is an adventurous and challenging proposal due to the following reasons: (1) The PIs based in have various domains of expertise, from theory of complex networks and nonlinear dynamics, to applied statistics in domains such as reliability and risk assessment, and development and application of operational research and operations management methods to MSN management and control problems. However, our expertise is complementary and will add a substantial body of new knowledge and bring novelties to the theory of complex networks, network dynamics and Bayesian networks, but also, applications of these new models to real-world MSN problems will ultimately lead to better understanding of complex MSN behaviour and will improve MSN management and control in the presence of risks and uncertainties. (2) This proposal will bring together PIs and PDRAs from 4 universities. The management of the resources involved is a challenge on its own. However, we believe that a very carefully designed project management plan can lead this research collaboration to its success. Furthermore, if funded, this research project can potentially secure the continuation of the collaboration among the four universities. (3) The project will involve a wide array of industrial partners from manufacturing primes (e.g. in Aerospace and Defence) to manufacturing trade organisations and consultants, to representatives of a brand new industry (offshore renewable energy) for which the in-bound MSNn does not yet exist.

Supply networks in three principal application domain, represented by our industrial partners, will be targeted for impact:
- Producer-driven inbound MSNs for industrial primes such as in the UK aerospace and communications industries.
- Global MSNs for key product categories that are critical within inbound MSN for many high-value manufacturing companies (e.g. titanium or composite pre-preg materials).
- Emerging UK industries such as renewable energy where MSN are unestablished, immature or evolving.

The main stakeholders impacted by this research are: manufacturing policy-makers and industry bodies; manufacturing businesses (including those with mature, global MSN such as aerospace and young sectors such as renewables); the professional communities of supply chain, risk management, mathematical science practitioners and analysts; the partner universities' students and wider research bodies (especially in related areas such as energy, manufacturing, management).

We seek to engage users via multiple mechanisms that will adapt to the different stages of the project. For those companies with whom we have already initiated a relationship and gained feedback that has shaped our proposal, we plan to continue bi-lateral dialogues between key researchers and company stakeholders (as named in our letters of support from e.g. Rolls Royce, PA Consulting, Tricorn Group, Nautricity, DSTL). Through these partnerships, we seek to gain an in-depth understanding of their MSN and develop longitudinal studies that should facilitate transfer of established knowledge to the company during the project, while informing our research path especially in relation to the development and validation of the modelling suite and the definition of meaningful performance measures.

Collaboration with industry bodies (e.g. National Composites Centre, RegenSW - see letters of support) and existing industry networks (e.g. Strathclyde Risk Consortium) provides access to a wide range of companies creating an opportunity to widen our industry involvement and impact. As we refine our research requirements for our empirical studies, we can draw upon the sample frame effectively provided by company members of such industry groupings. Our dissemination plan would include communications in form of practitioner articles, seminars and linked web resources to such industry groups so that we communicate research findings appropriately.