BT-based Orchestration
Mission logic encoded as modular Behavior Trees for deterministic sequencing and clear runtime supervision.
System / BT Diagram
Mission-level orchestration for an autonomous counter-UAV interceptor using Behavior Trees in ROS.
Status: Active Development — validated in simulation
At a Glance Features Architecture Services Testing Gallery Status
This project implements a mission manager for an autonomous UAV interceptor scenario, where the mission is executed as a set of Behavior Trees (BTs) ticking at runtime in ROS. The goal is to ensure a deterministic, testable and reactive mission flow for safety-critical operations (takeoff, navigation, interception, recovery, landing).
A key design goal is hybrid operation: the same mission can run fully autonomous, fully manual, or in a mixed mode, where an operator can trigger/override specific phases through mission-level services without restarting the system.
Mission logic encoded as modular Behavior Trees for deterministic sequencing and clear runtime supervision.
Autonomous, manual, or mixed execution via mission-level services (operator triggers/overrides at runtime).
Designed for repeatable test campaigns with structured logging, clear phase transitions, and failure handling.
The mission is decomposed into dedicated sub-trees (e.g., takeoff, pre-intercept, interceptor supervision, ending), coordinated by a top-level tree that gates and sequences the overall flow.
BT action/condition nodes interface with ROS topics/services and system state, enabling runtime checks (readiness, mode, completion conditions) and safe transitions.
Includes recovery behaviors and guard conditions to handle common failure modes (lost target, invalid state, interrupted mission steps, landing completion criteria).
The mission manager exposes a set of mission-level services that allow an operator to trigger or override key steps at runtime (useful for manual runs, debugging, and staged validation).
*Service names and exact interfaces are project-specific and may differ between deployments.*
The project has been validated through repeatable simulation campaigns with mission scenarios focusing on correctness, safety, and robustness of the mission flow.
Core mission flow: implemented and validated in simulation.
Ongoing work: improved flexibility for manual triggering, expanded failure handling, and interface refinement.
~75% — stable baseline complete, ongoing robustness and usability improvements.
BT diagrams and runtime logs (example snapshots).
Developed as part of the EAGLE.ONE ecosystem, leveraging ROS and Behavior Tree design patterns commonly used in robotics for reactive mission control and verifiable autonomy.