Enhance interoperability of autonomous systems.
Describe equivalent behaviors between autonomous frameworks.
Iterative development across the spectrum of autonomy.
AutonomyML ensures that different independent autonomous UxS system implementations publish their autonomy capabilities openly via a standard that enables the description of equivalent observable behavior, without relying on undocumented conventions inside its components.
AutonomyML relies on the OGC API - Connected Systems Standard (CSAPI) and SensorML standards as well as formalisms from Finite State Machines (FSM) and Behavioral Trees (BT) to create executable models of autonomous capabilities.
| Concern | Mechanism | What it provides |
|---|---|---|
| Interaction with a particular system | OGC CSAPI | System-facing observations, commands, and command-status resources. |
| Computational functions and dataflow | SensorML | Typed inputs, outputs, parameters, methods, and hierarchical process composition. Its framework includes computational processes, sensors, and actuators—not just sensor metadata. |
| Modes and transitions | FSMs | Explicit state-oriented behavior management. |
| Reactive task composition | BTs | Hierarchical selection and sequencing of conditions and actions, with execution status propagated through the tree. |
| Executable algorithms | Process implementations and system implementations | The actual perception, estimation, planning, control, or other computations invoked through these models. SensorML describes their interfaces and methodology; an execution environment supplies support for those methods. |
Capabilities, typed interfaces, process/behavior composition, FSM/BT primitive states & events, lifecycle, timing, and resource coordination.
Supported SensorML methods, OGC CSAPI mappings, FSM/BT dialects, platform capabilities, and domain-specific meanings.
Declarative planning, rich belief models, online learning, dynamic behavior generation, multi-agent coordination, and assurance artifacts.