ENCORE will commission and deliver Feasibility Studies to unearth as yet hidden potential for cross‐field collaborations through establishing areas in which collaborations and the exchange of expertise would be most effective, and initiate preliminary cross‐cutting work in the areas so identified.
Taking a method oriented approach we will undertake or identify, specify and articulate the need for comparative analyses across domains, that would identify common challenges and opportunities to reduce risk and characterise resilience in CES; 
1] Definition (and creation via feasibility studies) of integrated models that usefully describe the interactions between the technical infrastructure and its social and cyber‐physical contexts relating to shocks and stresses that define the dimensions of risk; 
2] Methodological efforts, aimed largely at capturing the network characteristics, both technical and social, of the system of systems; and 
3] Explicit testing and evaluation of the research through programs of collaboration with practitioners and governmental organisations.
Examples of the potential for the application of this field that respond to the ENCORE thematic areas and which will be considered for inclusion in the feasibility studies representing key scientific challenges and industrially crucial issues include:
  • Predicting equipment failures and their consequences in critical infrastructure systems;
  • Developing a management heuristic that plays the same role as a "risk register", but addresses systemic resilience; 
  • Optimising the deployment of instrumentation required to manage cities and other CES effectively;
  • Increasing the resilience of interdependent digital systems;
  • Advancing models of cascading failure on networks such that they take account of node heterogeneity and in particular the different failure/recovery modes of different types of node. 
  • Improving the number of contexts in which CES can be deployed with replicable performance;
  • Decreasing the likelihood of human behavioural errors in operating CES;
  • Identifying the critical elements that constrain/define system performance most strongly; 
  • Extending system lifetimes and functionality; 
  • Mapping the relationship between complex system complexity and fragility; 
  • Characterising uncertainty and defining the inference process to transition from one phase to the other in the control of CES and in complex decision making processes.
Our areas of thematic focus will evolve through the shared ideas of our network. Our scope is expressed through the vision, aims and objectives of ENCORE. Evidence of how this fits is best expressed through the tangible problems we seek to address, which are embedded in our initial underpinning themes that express how the high‐level problems defined above will be grounded in our activities:
Resilience in network and process dynamics:
A key problem for engineers is defining the relationship between how a CES is designed, the behaviour that this design produces in normal use, and how this behaviour alters when the system is subjected to multiple, overlapping shocks and stresses and also in the context of dynamics such as growth and adaptation. Reductionist assumptions on the nature of uncertainty and its propagation through the system can lead to incorrect expectations. An example is that the form of a city may be represented by a static transport network that exhibits a characteristic pattern of mobility, but we lack understanding of how this network interacts with associated energy, ICT and user networks operationally and under conditions of nodal or sub‐system failure. This issue is compounded when considering multiple overlapping uncertainties.
Understanding the constraints and design to benefit from these:
High‐order CES such as cities or national infrastructure exhibit constraints    that are often not considered in the models used to predict their behaviour. Processes such as aggregation and self organisation create temporal and relational constraints. We need to understand how such developmental constraints relate to design and performance in order to benefit, rather than suffer, from such characteristics (Thompson 1917Kauffman 1993Goodwin 2001). This is also linked to the CES being constrained to exist in a low‐dimensional medium: space‐time. Consider a real world energy network is a high‐dimensional abstract structure projected onto two spatial dimensions and allowed to evolve in time. A transport network is also high‐dimensional in its connectivity, but the structure of this connectivity is constrained by the physical infrastructure that it inhabits. Increasingly we are beginning to better understand the role of spatial and temporal constraints in the behaviour and configuration of complex engineering systems. We need systems of systems tools to express these higher dimensional characteristics in order to explore and understand their inherent risks and resilience. In the same way, we are beginning to understand developmental constraints in the growth and morphology of cities (Batty 2009) through exploiting understanding of the behaviour of complex systems (May 1976Feigenbaum 1980), but we are still mainly focused on the physical properties of cities. 
Leveraging natural world examples of complex systems:
Significant opportunities to improve resilience and performance lie in the potential to    extend this understanding into the underpinning engineering systems and social structures. Our aim is to extract the design principles that define beneficial attributes and behaviours in naturally complex systems and use this to inform the design of socio‐technical systems. As an example Advanced Energy eco‐systems, often referred to as Smart Grid 3.0, will require the properties of self‐organisation, self‐repair, robustness and adaptation characteristic of naturally complex systems (Carvallo, Cooper 2011).
Managing uncertainty in Complex Engineering Systems:
A challenge for engineers exists in their ability to alter and control the behaviour of existing CES to have predictable performance under uncertainty. Whether this is reconfiguring an energy system to accept a dynamic mix of generation types or updating software in communication networks, the implementation of minor changes can have cascading multi‐scale impacts upon systemic behaviour and performance (Newman 2011). Over time multiple changes are a record or description of the emergence in such systems. We need effective tools to deal with uncertainty when the identification of all statistical properties of a system is not possible due to its complexity. The risk associated with the emergence and propagation of uncertainties through complex systems renders the understanding of the degrees of redundancy or resilience of a network false. New techniques are required to understanding the transitional characteristics to develop our ability to predict such failures (Scheffer 2009).
Understanding risks in coupled socio­-technical systems:
Research is beginning to link models of CES to models of decision making (Vespignani 2012) to provide a more comprehensive understanding of both the impacts of CES design and management on societies and the potential for societies to be affect of their surrounding infrastructure. For example, water and power networks are key to the resilience of societies, but each is vulnerable to risk and uncertainty on a number of levels. Exploring the mathematical approaches to such coupling of human and engineering systems is a ‘blue skies’ topic. Applications of particular interest will include understanding uncertainty that should feed through to improved decision making for infrastructure projects.
Understanding the bi-directional impact of human behaviour on CES:
Modelling within CES tends to focus on the ways in which the technical components of systems interact. These systems do not operate in a vacuum as they are developed to service societal needs. Understanding the ways in which systems impact employee/user behaviour, and the ways in which employee/user behaviour impacts systems is crucial in order to identify dynamic issues that affect CES resilience. Similarly, understanding the impact of systems on individuals, communities and society, as well as the impact of individual, community or societal preferences, responses and expectations, will inform the understanding of the effectiveness of CES. Currently, assumptions about the impact of employee public behaviour on the resilience and effectiveness of CES tends to be built upon anecdote or assumptions. Recent research into emergency response, and critical national infrastructure employee willingness or ability to report to work during extreme events, demonstrates that these assumptions are often inaccurate (Rogers & Pearce, 2013). Additionally, cyber-security research into employee behaviour demonstrates an innate ability to find workarounds when the system fails to meet the needs of the user (Sasse, 2014). ENCORE’s activities will create research opportunities targeted at generating an empirical evidence base related to the impact of CES on human behaviour and the impact of human behaviour on CES in order to improve and adapt our current understanding of the resilience of CES.
Outwith the thematic areas above we would like to encourage innovative approaches to understanding complex systems. Experimental tools allowing the investigation of emergent behaviour, evolutionary approaches to selectively pick resilient solutions and convergent algorithms to pull complex systems back to optimal behaviour when disturbed are all methods within the scope of our mission. It is even conceivable that learning algorithms can adapt to new environmental constraints and increase system performance beyond that designed in at the outset, i.e., we give the machine the ability to learn new behaviours but allow the freedom for the machine to decide what it learns.

Share the NEWS