Decision making in asset management and planning is faced with many challenges, including: how to devise strategies that consider the dynamic and interdependent relationships among multiple and heterogonous assets and resources in a holistic way, planning under conditions of uncertainty, accounting for diverse stakeholders’ goals and preferences, and considering the long-term effects of investment and resource allocation decisions. Answering these questions, and more, requires integrated decision analysis methods that have the capability to integrate:

1. multiple sources and lines of evidence (such as historical data sets, expert opinion, and predictive models) to provide relevant and useable decision-support information;
2. knowledge and information about various parts aspects of the system (i.e. asset types, dependency types, processes and activities, resources, functions, performance indicators) in order to maximize value for stakeholders;
3. management and investment decisions across the assets lifecycle, with the view of identifying cumulative and long term effects on performance; and
4. various sources (e.g. data) and types (e.g. demand variability) of uncertainties with the provision of providing sufficiently accurate information to inform risk analysis.

The expertise of the CSC modelling program lies in the development of integrated decision analysis methods for asset management and planning. The program brings researchers from various and complimentary disciplinary backgrounds, who apply theories and methods to domains such as project management, supply chain design, life cycle analysis, risk analysis, workforce planning. Our decision analysis expertise cover areas such as predictive modelling, simulation-optimization frameworks, and multi-criteria decision analysis.

Decision making does not grow in a laboratory, but transpires in the real world. Therefore, our research focus is on developing and testing our novel multi-method decision analysis approaches using real-life problems and case studies. One of the main benefits is that direct interaction with real-life problems and authentic decision makers provides a powerful way to observe, interact and consequently understand the nature of the decision-making problems and decision maker’s information needs. This understanding ensures that our research is more accessible and relevant to end users, and maximizes our impact. The knowledge we accumulate through case studies is translated into domain lessons and underpins the agility of our development processes. We focus on developing the following methods and tools, with a particular focus on linking output to accessible decision insights in the application areas of interest.

Methodological Focus

Simulation Models. Simulation models are powerful predictive methods. The development of domain-specific modelling libraries and plug-and-play modelling technologies is still not accessible for decision makers. We focus on the methodologies and practices needed to develop reusable, extensible and user-friendly modelling libraries to capture domain knowledge and empower end users to take full advantage of simulation capabilities.

Optimization Engines. Optimisation principles can provide an alternative with the most cost-effective and performance-driven way under a set of prescribed constraints, by maximising desired factors and minimising undesired ones. Lack of real-life data and full information always impedes practice of optimisation techniques, which could be tackled by adopting advanced simulation models for business problems. Hence, integration of optimisation principles with simulation models (i.e., simulation optimisation) and multi-layer with multi-level optimisation is an important research area.

Multi-criteria Decision Analysis (MCDA). MCDA is an umbrella term covering decision analysis methods used to deal with problems involving multiple and complex criteria. MCDA is used in group decision making process to help the group elicit their goals in a form of a hierarchical tree of criteria and sub-criteria, and rank their preferences about a particular option in light of their priorities. A key strength of MCDA is the ability to handle complex problem with high dimensionality of criteria (e.g. economic, technical, environmental, security).

Design Structure Matrix (DSM). DSM supports the management of complexity by focusing attention on the interactions between components of a complex system (which can be products, organizations or processes). DSM-based techniques such as DSM clustering or partitioning have proven to be very efficient in understanding, improving and optimizing complex system architectures. A typical application example is product modular design. The ongoing research is to develop methodologies to manage the interdependencies in a system, to explore the new application of the available methods, and to tailor research and tool development to the emerging needs of practice.

Modelling and Exploration for Tradespace Multi-Attribute Tradespace exploration (MATE) is a conceptual design methodology that applies decision theory to modelling and simulation-based design to understand and assess alternative architectures. This is a linked method for progressing from vague user needs to preliminary design, supporting decision making in the preliminary design review (PDR) phase. Nowadays, non-traditional design criteria (referred as the “ilities”) are recognized as critical attributes of the tradespace, thus more work are required on developing proper models for defining, quantifying and visualizing these ilities.

Multi-method Decision Analysis. There is a recognition that the nature of complex problem necessitate the use of multiple decision analysis method to address different aspects of the problem. Yet, there is still limited knowledge into how these methods can be combined in practice to provide coherent decision insights.

Application Focus

Our application areas focuses on asset planning and management. For example:

Fleet Design and Management. Fleet management and design comprise activities that result in the optimal management of all aspects of a fleet throughout its life from the acquisition to the retirement. This includes fleet size, mix and deployment decisions together with inventories, supply chains, and other support systems.  During the life-cycle of a fleet, several resources are used and shared by assets of the fleet.  The task of fleet sharing resources among different entities (e.g. spatial regions, assets) and activities (e.g. operations and training) creates tightly coupled relationships among assets and their constituent systems. Reductionist thinking that fails to account for these interdependencies may lead to suboptimal outcomes of improving the performance of one asset or a function on the expense of another.

Capacity Planning. Asset planners are faced with the challenge to make decisions now whose outcomes can vary at different temporal scales and performance dimensions.  For example, a capacity investment containing more assets achieves a higher operational availability at the expense of higher capital and operational costs. On the other hand, keeping an existing asset and extending the life-span of assets may lead to a reduction in capital expenditure. However, this strategy may result in increased operating costs (i.e., more regular maintenance), and more importantly, it may harm the level of operational readiness (due to the increased probabilities of break-downs). How can simulation-optimization frameworks help decisions regarding timing and amount of investment to optimize outcomes and achieve those trade-offs?

Workforce Planning. The design and management of complex workforce systems remain a challenge because of the multiple delays and feedback connections. For example, the pressure to cut recruitment rates will lead to a short-term reduction in costs, but can have detrimental consequences on sustainability of training pipelines and the availability of future workforce.  How can the use of simulation models help predict these flow-on effects? How can we make use of decision analysis to build and examine the effects of different workforce designs and training pipelines?

Project Management and Scheduling. The project schedule in project management reflects all the work associated with delivering the project on-time, which communicates what tasks need to be done along with their resource usage and execution time. Scheduling multiple projects under scarce resource conditions is often treated as a complex problem and combinatorial in nature (in optimisation context). Hence, implementing OR principles in this highly dynamic and complex domain has always been considered as a challenging research problem. (i) How can we apply surrogate optimisation techniques in this discrete domain (i.e., resource constrained project scheduling)? (ii) How to implement bi-level and multi-layer optimisation concepts in supply chain integrated project scheduling context? (iii) How to maintain sustainability in cloud projects are some of our future research directions?   

ResIlience and Risk Planning. Many important planning problems involve decision making under uncertainty—that is, choosing actions based on often imperfect observations, with unknown outcomes.  Choosing actions based on the assumption that average conditions will occur are usually wrong, as stated by “The Flaw of Averages.”  Once asset planners have developed alternative strategies. Decision support methodologies that use the concept of robustness can help address this challenge, by identifying actions and resilient plans that perform well across many different deeply uncertain future conditions and minimize the risks of falling below the acceptable threshold values of key performance indicators.