"-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"> NASA :: Intelligent Systems :: APEX :: Projects :: Autonomous Rotorcraft Project
Skip Navigation
Home | Organization | News/Events | Research | Publications | Destinations | Partnering | Software
Autonomous Systems and Robotics
Control Agent Architectures
Apex
Apex Overview | Projects | Software | Documentation | Publications | People | Contact Us
Autonomous Rotorcraft Project
Autonomous Rotorcraft Project Mission Planner Interface with RMAX helicopter inset

The Autonomous Rotorcraft Project is a NASA/Army effort to produce an intelligent, highly flexible aerial observation system. Users interact with the system by expressing information needs and preferences rather than by commanding the vehicle directly. For instance, a user might specify an interest in knowing as soon as possible if a fire has broken out at a particular location.

Apex: Deliberate, Control, Monitor and Interpret. Apex, the intelligent behavior component of the Autonomous Rotorcraft Project, generates mission plans to best meet collections of such requests. In doing so, Apex takes into account factors such as how probable events are, projected plan costs, and the value of observation targets. Apex then executes these mission plans in conformance with standard operating procedures.

Apex's execution capabilities provide high-level task control including plan refinement; monitoring and interpreting off-nominal conditions and other contingencies (including user-initiated changes in mission plans and constraints); and coordinating subsystem resources and activities, for example, at-target maneuvers, attitude control, obstacle avoidance path planning and trajectory determination.

Platform

A Yamaha RMAX helicopter was selected as a platform to demonstrate Autonomous Rotorcraft Project technologies. The RMAX is an unmanned subscale vehicle (3m rotor diameter) with a one hour hovering flight duration, 40 knot maximum airspeed and 65 lb payload capacity. These characteristics represent a sweet spot in the tradeoff between capability and cost in that they allow us to fly diverse, sophisticated technologies onboard the aircraft and to fly far enough and for long enough to carry out complex, interesting missions. We have modified the base platform to include an enhanced avionics package (supporting autonomous takeoff, landing, navigation, communication and computing) and a sensor payload that includes video cameras a stereo camera pair and a laser range finder useful for both obstacle avoidance and for providing users with 3D images of the environment.

Capabilities

MPI

Mission Planning Interface (MPI)
The Mission Planning Interface allows an operator to define a mission for the rotorcraft using a pan and zoom map interface. Live Demo is now available. (Details...)


   Mission Plan Generation

Mission Plan Generation
The Mission Plan Generator addresses the challenge of maintaining awareness of a large number of sites with a small number of UAV (unmanned aerial vechicles). (Details...)
Passive Sensing Image

Passive obstacle sensing and mapping
A range map is generated every five seconds using a stereo pipeline. A wide baseline gives accurate disparity at long distances. (Details...)


   Active Sensing

Active obstacle sensing and mapping
Active sensing with a 81 meter range and 1 cm accuracy is achieved usin a SICK PLS scanning laser with a 180 degree field-of-view. The laser has 0.5 degree resolution at 37.5 Hz. (Details...)
Maneuvers

Maneuvers
The Apex team has developed maneuvers for use with the Autonomous Rotorcraft. These maneuvers allow the rotorcraft to approach and observe a target in a variety of ways. ( Details... )


   Flight control

Flight control system
Model-following control law provides altitude stablization and waypoint guidance. (Details...)
Condition monitoring

Visualizing Autonomy Logic
Sherpa helps you debug Apex applications. It provides a number of ways of examing the state of your application to better understand what your code is doing. (Details...)


  Simulation

Desktop and hardware in-the-loop simulation
Integrated math models include validated hover/low speed and forward-flight linear models identified from flight data. (Details...)

Demonstration videos

ARP Five Year Demo Autonomous Rotorcraft Project Five Year Demonstration In this flight, Apex mission-level autonomy capabilities were demonstrated. For more information, see the ARP Mission-level Autonomy presentation.

Streaming video in QuickTime MOV file format

Download video in MOV file format
ARP overview Autonomous Rotorcraft Project Overview This video provides a general overview of the Automous Rotocraft Project.

Streaming video in QuickTime MOV file format

Download video in MOV file format
Hunter Liggett Flight Fort Hunter Liggett Flight In this flight, the rotorcraft avoided real obstacles (trees) that were identified by scanning laser.

Streaming video in QuickTime MOV file format

Download video in MOV file format
August 18, 2005 Flight Flight - August 18, 2005 In this flight, the rotocraft executes an area spiral.

Streaming video in QuickTime MOV file format

Download video in MOV file format
Flight around collapsed building Final rehearsal of DART exercise This rehearsal included flying around a collapsed building structure, observing multiple targets, and viewing targets from different headings.

Streaming video in QuickTime MOV file format

Download video in MOV file format
DART test Another DART test flight around a collapsed building.

Streaming video in QuickTime MOV file format

Download video in MOV file format
Flight test Obstacle avoidance Flight test with obstacle avoidance path planning.

Streaming video in QuickTime MOV file format

Download video in MOV file format

Autonomous Rotocraft Team

Autonomous Rotorcraft Project team

The Autonomous Rotocraft Project lead is Matt Whalley. Michael Freed is the Apex team lead.

Team members pictured (left to right) are Marc Takehashi, Robert Harris (Apex team), Matt Whalley (project lead), Michael Freed (Apex team lead), Greg Schulein, Perry Kavros. Not pictured: Will Fitzgerald (Apex team).