Autonomous & Resilient Countermeasures for Emergent System Disruptions with Application to Air Traffic Management Open Access
Downloadable ContentDownload PDF
Engineered systems are susceptible to the disruption of intended functionality when unanticipated operating environments and constraints emerge during mission execution. To safeguard intended system functionality, resilient systems which integrate disruption avoidance and mitigation measures are needed. System automation methods have been successfully adopted for disruption avoidance and performance optimization based on anticipated needs and risks. However, given the development of unpredicted or emergent issues, human intervention is often relied on as an operational contingency to system failure or degraded performance. The purpose of this research was to develop and analyze a model for evaluating and selecting system responses to mitigate emergent disruptions without human intervention. The scope of this research addresses risk management within the system lifecycle with focus on system adaptation to unpredictable changes in the operational environment or functional constraints. The Disruption Resilience and Adaptation Model (DREAM) was formulated as an autonomous decision-making and performance evaluation process for mitigating unforeseen or unavoidable system disruptions. Evaluating the system as a whole, the model integrates diagnostic and prognostic heuristics for performance feedback and regulating system responses to actual or potential disruptions. The suitability of rule and utility-based automation methods as system adaptation techniques was also investigated.Via modeling and simulation, the DREAM and existing system automation methods were applied to an Air Traffic Management (ATM) problem and compared based on: (i) mission reliability; (ii) incident rate; (iii) system efficiency; (iv) standard deviation of efficiency; (v) and system stability. Simulation results demonstrate the DREAM yielded statistically significant reductions in the frequency and severity of ATM system disruptions in comparison to the existing rule-based standard and the leading utility-based method. As demonstrated in the ATM case study performed in this research, the DREAM can reduce the frequency and severity of system disruptions that are unanticipated or unpredictable. Furthermore, the DREAM can autonomously enable preventive and corrective action in response to unexpected system disruptions, increasing the likelihood of achieving intended system objectives without the need for human intervention. In practical terms, use of the DREAM could increase the operational availability of safety- and security-critical systems, such as ATM systems, for which disruptions can be catastrophic in nature. Beyond the ATM application in this research, the DREAM is targeted towards engineered systems which are susceptible to inevitable, yet unavoidable disruptions, such as natural disasters, intentional attacks, and human errors.