Software Engineering Institute | Carnegie Mellon University
Software Engineering Institute | Carnegie Mellon University

Collaborative Autonomy

Collaborative Autonomy for Mobile Systems architects, designs, analyzes, and validates portable architecture and middleware to support user-directed groups of autonomous sensors and systems. The current focus is on
  • middleware that creates a decentralized, distributed operating environment for swarms of sensors and robots, guided by a human user
  • area coverage techniques that specialize in prioritized zones and mission objectives
  • algorithms that prioritize information flows and route mobile sensors/drones/robots into locations that best serve mission utility


Collaborative Autonomy Challenges

  • Autonomy focus is on single unit control.
  • Focus is on centralized controllers (prone to failure/attack).
  • Autonomy frameworks tend to be targeted at homogeneous platforms and algorithms.
  • Blocking communications are prone to faults/attacks/ outages/loss-of-control. GPS is highly inaccurate for precise maneuvers.
  • There is a lack of standardization for autonomous collaboration.
Illustrates GAMS platform and algorithm interactions

Our Approach to Collaborative Autonomy

  1. Create a portable, open-sourced, decentralized operating environment for autonomous control and feedback. Focus on scalability, performance, and extensibility.
  2. Integrate the operating environment into unmanned autonomous systems (UAS), platforms, smartphones, tablets, and other devices. Focus on portability.
  3. Design algorithms and tools to perform mission-oriented tasks such as area coverage and network bridging between squads.
  4. Design user interfaces to help single human operators control and understand a swarm of UAS, devices, and sensors (human-in-the-loop autonomy).