Is this project an undergraduate, graduate, or faculty project?
Graduate
Project Type
individual
Campus
Daytona Beach
Authors' Class Standing
Fayruz Maysha, Graduate Student
Lead Presenter's Name
Fayruz Maysha
Lead Presenter's College
DB College of Engineering
Faculty Mentor Name
Bryan Watson
Abstract
Aerospace autonomous systems and multi-agent systems (MAS) are increasingly deployed in safety-critical operations such as wildfire response, disaster management, and distributed sensing. These systems rely on coordinated interactions among heterogeneous agents operating under uncertain and failure-prone conditions. However, existing research on resilience in MAS is predominantly algorithm-centric, focusing on control strategies, learning adaptation, and communication robustness while largely assuming stable system structures and reliable operating conditions. As a result, there is limited understanding of how infrastructure composition itself influences resilience under execution-time failure. This study addresses this gap by introducing an infrastructure-level perspective on resilience through the Kinship Infrastructure Design (KID) framework. In this approach, system components are modeled as agents with fixed functional traits, and system composition is quantified using a kinship coefficient (Φ) that captures the degree of functional similarity and overlap among agents. A simulation-based experimental methodology is employed, modeling a wildfire-response aerospace architecture as a heterogeneous multi-agent system. Multiple infrastructure configurations are generated under a fixed mission budget and identical objectives, and are subjected to controlled execution-time disturbances, including communication degradation, sensing disruptions, and partial agent failure. System performance is evaluated using resilience-oriented metrics such as coordination persistence, graceful degradation, and recovery behavior. The study systematically analyzes how varying levels of kinship influence system performance under degraded conditions. The central hypothesis is that moderate kinship configurations achieve superior resilience, balancing redundancy and diversity to enable sustained coordination without introducing excessive fragility or inefficiency. The expected contributions of this work are threefold. First, it provides a formal framework for representing and quantifying infrastructure composition in aerospace systems. Second, it establishes a systematic methodology for evaluating resilience as a function of design-time decisions rather than solely algorithmic adaptation. Third, it generates actionable design insights linking functional diversity and overlap to resilience outcomes in complex, failure-prone environments. By shifting the focus from reactive, algorithm-level solutions to proactive, infrastructure-level design, this research advances the understanding of resilience in aerospace autonomous systems and supports the development of architectures that maintain functionality under real-world operational disruptions.
Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?
No
Designing Resilience: A Kinship-Based Framework for Infrastructure Composition in Aerospace Autonomous Systems
Aerospace autonomous systems and multi-agent systems (MAS) are increasingly deployed in safety-critical operations such as wildfire response, disaster management, and distributed sensing. These systems rely on coordinated interactions among heterogeneous agents operating under uncertain and failure-prone conditions. However, existing research on resilience in MAS is predominantly algorithm-centric, focusing on control strategies, learning adaptation, and communication robustness while largely assuming stable system structures and reliable operating conditions. As a result, there is limited understanding of how infrastructure composition itself influences resilience under execution-time failure. This study addresses this gap by introducing an infrastructure-level perspective on resilience through the Kinship Infrastructure Design (KID) framework. In this approach, system components are modeled as agents with fixed functional traits, and system composition is quantified using a kinship coefficient (Φ) that captures the degree of functional similarity and overlap among agents. A simulation-based experimental methodology is employed, modeling a wildfire-response aerospace architecture as a heterogeneous multi-agent system. Multiple infrastructure configurations are generated under a fixed mission budget and identical objectives, and are subjected to controlled execution-time disturbances, including communication degradation, sensing disruptions, and partial agent failure. System performance is evaluated using resilience-oriented metrics such as coordination persistence, graceful degradation, and recovery behavior. The study systematically analyzes how varying levels of kinship influence system performance under degraded conditions. The central hypothesis is that moderate kinship configurations achieve superior resilience, balancing redundancy and diversity to enable sustained coordination without introducing excessive fragility or inefficiency. The expected contributions of this work are threefold. First, it provides a formal framework for representing and quantifying infrastructure composition in aerospace systems. Second, it establishes a systematic methodology for evaluating resilience as a function of design-time decisions rather than solely algorithmic adaptation. Third, it generates actionable design insights linking functional diversity and overlap to resilience outcomes in complex, failure-prone environments. By shifting the focus from reactive, algorithm-level solutions to proactive, infrastructure-level design, this research advances the understanding of resilience in aerospace autonomous systems and supports the development of architectures that maintain functionality under real-world operational disruptions.