Applying model-based systems engineering to the development of a test and evaluation tool for unmanned autonomous systems

In this paper, we apply the Model-Based Systems Engineering (MBSE) concepts and approaches to the early phases of the development of a Test & Evaluation (T&E) tool for Unmanned Autonomous Systems (UASs). This helps meet the design requirements and maintain traceability (of design requirements and decisions for satisfying stakeholder's needs). UAS development is driving toward increasing levels of autonomy for unmanned systems. The dynamic, non-deterministic behavior of intelligent autonomous systems presents the testers with a significant challenge. The ability to predict the behavior and evaluate performance of increasingly intelligent systems, especially those that employ vision-based behaviors, is seen as a critical T&E shortfall. To address this challenge, we propose, in this paper, to use a high-fidelity simulation environment. This can significantly aid in the evaluation of UAS behaviors and their perception mechanisms. Such a high-fidelity simulator enables the testes to safely conduct a wide variety of mission scenarios to test an autonomous system by providing truth data to compare with the UAS's perceptions. A major challenge here is to manage the system modeling complexity and maintain traceability of design decisions made at each level of the development to meet stakeholder's needs. In this paper, we follow MBSE methodology and use Systems Modeling Language (SysML - a domain-specific modeling language for systems engineering used to specify, analyze, design, optimize, and verify systems) to establish a systematic framework for designing a T&E tool for UASs and to transform stakeholder's needs into design requirements to maintain traceability.