Vehicle Intelligence


Research examines the vehicle's "understanding" of its own state and that of its local environment, in the context of its mission, background knowledge, and organic sensor information. It includes research on specialized sensors and concomitant processing; distributed perception to enable navigation and manipulation of extremely small platforms; and vehicle planning and monitoring behaviors supported through combined sensory and contextual information in a world model.

  • Sensing & Sensor Processing
  • Semantic Perception (Scene Understanding)
  • Distributed Perception/Fusion

Intelligence and Control

Research examines vehicle behaviors, including planning, monitoring, and correcting behaviors to achieve desired mission goals. It focuses heavily on mechanisms for learning, both supervised and unsupervised; for continual or life-long learning, and for generalization. Research in this area focuses on effective mechanisms for creating increasingly complex and adaptive behaviors from elemental machine skills, capabilities that will enable the vehicle to effectively team with Soldiers and other unmanned vehicles at the operational tempo of the mission. These efforts include the means for creating behaviors for individual vehicles, as well behaviors for groups of homogeneous or heterogeneous vehicles working together to achieve a singular goal. We generally describe the vehicle control architecture as hierarchical and possessing three layers: a low level control layer, a vehicle centric middle layer, and a global mission level layer.

  • Control
  • Planning/Guidance
  • Abstract Reasoning
  • Teaming & Coordination
  • Behaviors
  • Learning & Adaptation

Human-Robot Interaction (HRI)

Research focuses on interactions between humans and robots/intelligent platforms. It examines mechanisms for effective robot communication between Soldiers and robots; communication or transmittal of information in its most elemental sense – including the understanding of gesture and voice; the use of language as an abstraction for transmitting information; and intra-team behavior. Long term research also includes efforts to measure the human state and explore new architectures that incorporate insight into the operator state and intention in Human/Autonomous System Decision Architectures for the integration of human adaptive abilities for enhanced autonomy in complex and dynamic conditions. Finally, it provides models of societal interaction to enable construction of robot behaviors that will lie within societal behavior norms and create a common model of behavior.

  • Soldier-Machine Communication
  • Intra-Team Behavior
  • Societal Interaction
  • Intra-Team Behavior
  • Societal Interaction

Last Update / Reviewed: February 5, 2015