Prev-Talk | Next-Talk | All-Talks | Talks-Sorted

Talk-10: An Overview of Novel Technologies Leveraging MOOS-IvP Developed at the Naval Surface Warfare Center, Panama City Division

Matthew Bays, Naval Surface Warfare Center, Panama City Division

We present an overview of various autonomy-related technologies currently under development at the Naval Surface Warfare Center, Panama City Division. We believe the technologies discussed have the potential to significantly increase the capabilities of MOOS-IvP by expanding the use of MOOS-IvP to task-level control paradigms, allow for reduced difficulty in autonomy deployment, and , further integrate the IvP-centric capabilities to the Robot Operating System.

The first technology we discuss is a task-level autonomy framework. While autonomy developers have been able to take advantage of external control interfaces on intelligent vehicles for years, they have typically had to choose between control architectures and interfaces that only allowed them to use the existing behaviors of the vehicle (task-level control) or those that allowed them to set desired control values (setpoint control) but then required them to re-invent and re-tune behaviors that might already be part of the vehicle’s native programming. As external control interfaces become increasingly open and allow developers access to both types of control, a control architecture is necessary that takes advantage of the strengths of each method and can switch between them as appropriate. We present one solution to this challenge that not only accomplishes these goals but is also vehicle agnostic and extendable to an arbitrary number of low-level controllers. The solution allows in-situ use of internal vehicle behaviors while providing the flexibility of backseat-driver control using MOOS-IvP.

The second technology of which we will provide an overview is the Autonomy in a Box container environment. Autonomy in a Box is a method to dramatically reduce the level of effort and lead time required to take autonomy algorithms from initial development to field experiments when using shared assets. The method leverages the Docker containerization environment coupled with automatic configuration and deployment modules and the MOOS-IvP autonomy framework. The result is a quickly-deployable, easily reconfigurable, and vehicle-agnostic autonomy solution for use when assets are shared and repeatedly re-baselining the system is necessary. For an initial implementation of the containerization system, we leverage the Mission-Oriented Operating Suite – Interval Programming autonomy framework.

Finally, we present a set of software packages that integrates the arbitration framework, user interface utilities, and native vehicle behaviors found in the Mission-Oriented Operating Suite (MOOS-IvP) into the Robot Operating System. MOOS-IvP is prevalent in the maritime robotics community due to a large library of maritime-specific autonomy behaviors, interfaces to common unmanned underwater vehicles, utility programs, and the inclusion of a default arbitration framework for autonomous decision making. However, the Robot Operating System (ROS) has grown in use within the autonomy and robotics communities at large due to the utility of its communications framework, node introspection capabilities, and messaging architecture. In our work we combine the best capabilities of both robotics architectures into two packages for use in maritime autonomy applications. The packages allow introspection of the various components of the IvP autonomy framework and allow behaviors to be backwards-compatible with standard MOOS-IvP.

Categories:

  • AUVs
  • Simulation/Visualization
  • Navigation/SLAM
  • Payload Autonomy Platform/Interface