CASoS Engineering Design Process

A general CASoS Engineering design process must be wide and deep to cover the many potential opportunities for unexpected, nonlinear, interconnected behavior and to find and make use of similarities across many disciplines. As illustrated below, the design process is comprised of three phases applied primarily in succession but with some overlap, blending and iteration (both between and within) to deliver CASoS engineered solutions.

• Defining (blackboard to details) 1) the CASoS of interest, 2) the Aspirations (Predict, Prevent or Cause, Prepare, Monitor, Control, and Recover or Change), 3) Appropriate conceptual models (including required data) and 4) Testing to understand and compare aspirations. Possible methods, theories and fields of contribution include analogy, dimensional analysis and similitude, experimental design, system dynamics, non-equilibrium thermodynamics, complex adaptive systems, game theory, percolation phenomena, agent-based modeling, networks, system optimization and control, and many others. A taxonomy of problems and approaches to their solutions is unfolding from this activity.

• Designing and Testing Solutions using computational models, data mining/integration, experiments, etc, within a common quantitative analysis environment. Designing and testing solutions are problem dependent, focused on answering questions relevant to any aspiration: 1) how might the aspiration be implemented and are there feasible choices within the multi-objective space, 2) how robust are these choices to uncertainties in assumptions and 3) what are the critical enablers that increase system resilience. Included in this process is the delineation of unintended consequences and their amelioration/mitigation. Integral to this process are Uncertainty Quantification and Verification and Validation efforts.

• Actualizing the engineered solutions devised through application within the real system. The engineered solution may be a concept, a computational tool, a sensor, a control policy or a CASoS itself. This activity involves working with decision makers (change the world), other researchers (change the field) and people affected by the change (understand the impact). This involvement requires a long term commitment: these are high-consequence systems that adapt to change. Any change makes us part of the system with concurrent obligation through a solution’s lifecycle. Monitoring of system response is required, which may trigger the redefinition, redesign and retesting of solutions if found not to work.

Actualization will require designing and testing to suggest future steps: adaptations of the system might require us to return to fundamental thinking to understand the adaptations and possibly redefine the CASoS and our aspirations as the system changes. The current structure and state of the CASoS Engineering Design Process are provided in Complex Adaptive Systems of Systems (CASoS) Engineering Framework Version 1.0.