The concept of resilience has emerged from a number of domains to address how systems, people as well as organisations can handle uncertainty and thereby not only survive hardship, but also thrive and prosper. This is of particular importance for engineering infrastructure systems which, due to the inherently long lifecycles giving rise to many unknowns, need to be designed for resilience such that it not only maintains operations in the face of day-to-day demands, but also continue to be able to evolve for the future. While there has been substantial interest in resilience from both academia and industry, exactly how such systems may be endowed with resilience to address these concerns from an engineering design perspective is less clear.

To this end, a literature review was first conducted to compile the definitions and characteristics of resilience across the domains of engineering, organisational management and ecology. The characteristics were found to comprise: absorbing disturbances, adapting for change and thriving for the future. These were then mapped to the engineering design ilities of robustness, adaptability and flexibility before being brought together in a conceptual model to form a strategic view for resilience. Further methods from resilience and engineering design literature were then consulted to understand how this particular view could be modelled and evaluated. This led to the development of a preliminary model using the Least Squares Monte Carlo method adapted for a telecommunications case study.

The insights gained from these explorations were then used to drive the synthesis of a novel support method whereby the design for flexibility framework was adapted to include decision modelling with Bayesian Networks and for resilience analysis. Here, resilience is taken to be the maximisation of the system economic lifecycle value under uncertainty, as measured by Expected Net Present Value, through robust and flexible strategies. This was applied to two case studies involving infrastructure systems: the first built upon existing work based on a Waste-to-Energy system in Singapore to verify the new method while the second applied the support method with BT, a multinational telecommunications company based in the UK, to gauge reception of this approach in industry. In both cases, the initial capacity and maximum number of upgrades served as proxies for robustness and flexibility respectively. Results demonstrate that Bayesian Networks are able to model decision rules for flexibility by selecting technology options over time given observations on the system and are also useful for extracting expert domain knowledge. While the construction of Bayesian Networks are subjective, they present an intuitive visualisation of the dependencies in a system and as such, engaged stakeholder interest. Resilience analysis examined the effect of volatility and drift of demand on the design strategies and indeed, there existed a trade-off between robust and flexible strategies. Furthermore, the greater utility of the support method lies in aiding decision makers in exploring the solution space and prompting discussions for what-if scenarios for the organisation.