Science of Autonomy

What Is It?

The Science of Autonomy Program addresses critical multidisciplinary autonomy challenges that cut across different Office of Naval Research (ONR) departments and warfighting areas/domains including air, sea, undersea and ground.

How Does It Work?

This program develops collaborations among researchers in different autonomous system domains that have traditionally been somewhat separated, such as air, sea, undersea, ground, control theory, computational intelligence, human factors, biology, economics, cognitive science/psychology and neuroscience

What Will It Accomplish?

Autonmous operations—in many naval domains, for many mission types, and on multiple platforms—can operate as part of a hybrid force with manned systems and platforms and maintain survivability through decentralized assets/redundancy. They can also reduce the need to place personnel and high-value assets in high-threat areas, reduce manning and communications requirements, expand the operational envelope of naval forces, and provide force multiplication or replace existing capability with a less expensive alternatives.

The Science of Autonomy Program addresses technical challenges related to autonomy that were identified in a series of ONR and Naval Research Laboratory workshops. The challenges are in the four interrelated areas of human collaboration with autonomous systems, perception and intelligent decision-making, scalable distributed collaboration and intelligent architectures. These challenges need to be addressed relative to critical aspects of the naval domain, including:

  1. Operations in spatially and temporally variable and uncertain environments with limiting manning, communications, and other resources
  2. Users with a wide range of skills and experience, including unmanned system services that can support small tactical units
  3. Diverse environments encompassing air, sea surface, undersea and ground systems and hybrid concepts in between
  4. Platforms with highly limited and intermittent communications
  5. Complex missions with heterogeneous platforms and sensors, including significant differences in physical and sensing capabilities
  6. Rapid and dynamic responses to needs and changes in the operating space
  7. The need for automation to explain its capabilities to the user and reliably execute the required tasks in the required time.

Examples of multidisciplinary research include:

  1. A control engineer working with a neuroscientist to develop spatial understanding approaches for autonomous systems. These systems would fit human semantic models that can be used to create unmanned aerial “wingmen” for dismounted Marines
  2. Biologists and engineers using models of social interactions in animal groups that allow individuals to access higher-order computational abilities at the collective level and make good decisions despite uncertainty
  3. Biologists, psychologists and engineers applying behavioral and cognitive models of predator-prey relationships to engineered systems for intelligence, surveillance and reconnaissance, or ISR, of large, complex areas by
    heterogeneous unmanned systems.

Research Challenges and Opportunities:

  • Scalable, self-organizing, survivable, organizational structure/hierarchy of heterogeneous unmanned air, ground, surface and underwater vehicles and sensors appropriate to naval mission domains
  • Autonomous learning, reasoning and decision-making in unstructured, dynamic and uncertain environments
  • Human interaction and collaboration including understanding intent and actions of human team members, adversaries and bystanders
  • Organic perception and understanding to support decision-making, reasoning and actions in a complex, dynamic world

Point of Contact:

 Marc Steinberg
 (703) 696-5115
 marc.steinberg@navy.mil

* Some pages on this website provide links which require Adobe Reader to view.