Continuous Adaptive Runtime Integration Testbed for Complex and Autonomous Systems

Abstract

The integration of complex, distributed systems typically takes thousands of man-hours and years of detailed design and testing. Despite all this effort, the effectiveness of classical system integration becomes a major issue when one adds the runtime adaptive behavior that is increasingly vital for dynamic, autonomous systems influenced by their surroundings. Self-integrating systems proposed in recent research strive to autonomously integrate new components, reducing design and testing costs. However, these systems are challenging to validate, especially at scale. This paper describes CHARIOT, Continuous High-level Adaptive Runtime IntegratiOn Testbed, which allows for different approaches and systems to be dynamically deployed, assessed and compared on a shared common platform. CHARIOT uses self-flying drones and self-driving cars to validate autonomous signal integration. This paper will discuss our early work in designing the CHARIOT architecture and integration protocol.