Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Collaborative Technology Alliance(CTA)
Charles ShoemakerARL Collaborative Alliance Manager
Kevin BonnerConsortium Manager, General Dynamics Robotic Systems
Robotics
ObjectivesObjectivesConsortium PartnersConsortium Partners
GD Robotic Systems (Lead)
Applied Systems Intelligence
BAE Systems Jet Propulsion
Lab Micro Analysis &
Design Sarnoff
Corporation SRI International Carnegie Mellon
University Florida A&M
University University of
Maryland PercepTek SSC
Technical AreasTechnical Areas Perception
Intelligent Control & Behaviors
Human-Machine Interface
Make the research investments that support the Army’s autonomous mobility goals:
•Develop perception technologies that allow robotic vehicles to understand their environment;
•Develop intelligent control technologies enabling robotic systems to autonomously plan, execute, and monitor operational tasks undertaken in complex, tactical environments;
•Develop human-machine interfaces that allow soldiers to effectively task robotic systems and minimize operator workload.
RoboticsRoboticsCollaborative Technology Alliance Collaborative Technology Alliance
Execute
RoboticsRoboticsCollaborative Technology Collaborative Technology
AllianceAlliance CM: GDRS, Kevin Bonner CAM: ARL, Charles Shoemaker
RoboticsRoboticsCollaborative Technology Collaborative Technology
AllianceAlliance CM: GDRS, Kevin Bonner CAM: ARL, Charles Shoemaker
Intelligent Control Architecture
CMU, Dr. Tony Stentz ARL, Stuart Young
Intelligent Control Architecture
CMU, Dr. Tony Stentz ARL, Stuart Young
Advanced Perception
CMU, Dr. Martial Hebert ARL, Dr. Nasser
Nasrabadi
Advanced Perception
CMU, Dr. Martial Hebert ARL, Dr. Nasser
Nasrabadi
Human-Machine Interface
MAAD, Dave DahnARL, Dr. Rene dePontbriand
Human-Machine Interface
MAAD, Dave DahnARL, Dr. Rene dePontbriand
Terrain Characterizatio
n & Obstacle Detection
Terrain Characterizatio
n & Obstacle Detection
Perception for Mobility in Cluttered &
Dynamic Environments
Perception for Mobility in Cluttered &
Dynamic Environments
Mid-range Perception
Mid-range Perception
Architectures for Robotic
Battle Teams
Architectures for Robotic
Battle Teams
Planning Solutions
Planning Solutions
Tactical and Cooperative Behaviors
Tactical and Cooperative Behaviors
Fault Detection, Isolation &
Accommodation
Fault Detection, Isolation &
Accommodation
Scalable Soldier-Machine
Interfaces
Scalable Soldier-Machine
Interfaces
Human Performance Models and Workload
Assessments
Human Performance Models and Workload
Assessments
Multi-modal interactions
Multi-modal interactions
Enables robotic vehicles to understand
their environment
Perceive
Enables robotic
vehicles to intelligently
plan & execute military missions
Control
Enables soldiers to
seamlessly task robotic assets for missions as part of a mixed combat team
Supervise
RoboticsCollaborative Technology Alliance
Major Accomplishments
RoboticsCollaborative Technology Alliance
Major Accomplishments
In “Perception”
Advances in:
•LADAR processing refinements leading to finer resolution and better separation of the objects from their backgrounds
•New stereo techniques that are tuned to complex environments such as forests and grassy environments
•Detecting water
•Detecting and identifying thin wires
•Detecting moving objects
Results in faster, stealthier robots
In “Perception”
Advances in:
•LADAR processing refinements leading to finer resolution and better separation of the objects from their backgrounds
•New stereo techniques that are tuned to complex environments such as forests and grassy environments
•Detecting water
•Detecting and identifying thin wires
•Detecting moving objects
Results in faster, stealthier robots
RoboticsCollaborative Technology Alliance
Major Accomplishments
RoboticsCollaborative Technology Alliance
Major Accomplishments
In Intelligent Control
Advances in:
•Robust local planning
•Maneuver in dynamic environments
•Tactical behaviors
•Collaborative operations
Enables collaborative operation of manned & unmanned air & ground platforms
In Intelligent Control
Advances in:
•Robust local planning
•Maneuver in dynamic environments
•Tactical behaviors
•Collaborative operations
Enables collaborative operation of manned & unmanned air & ground platforms
RoboticsCollaborative Technology Alliance
Major Accomplishments
RoboticsCollaborative Technology Alliance
Major Accomplishments
In Intelligent Control
Advances in:
•Robust local planning
•Maneuver in dynamic environments
•Tactical behaviors
•Collaborative operations
Enables collaborative operation of manned & unmanned air & ground platforms
In Intelligent Control
Advances in:
•Robust local planning
•Maneuver in dynamic environments
•Tactical behaviors
•Collaborative operations
Enables collaborative operation of manned & unmanned air & ground platforms
RoboticsCollaborative Technology Alliance
Major Accomplishments
RoboticsCollaborative Technology Alliance
Major Accomplishments
Human-Machine Interfaces
Advances in:•Scalable Operator Control Unit•Multi-mode control•Spoken Language Interface
Reduces the workload of controlling multiple robotic assets
Human-Machine Interfaces
Advances in:•Scalable Operator Control Unit•Multi-mode control•Spoken Language Interface
Reduces the workload of controlling multiple robotic assets
OCU with Map view
OCU with RSTA view
RoboticsCollaborative Technology Alliance
Focused Initiatives
RoboticsCollaborative Technology Alliance
Focused Initiatives
Object Classification & Identification
Object Classification & Identification
•Multi-sensor registration & fusion:
•Scene segmentation
•Ability to detect & classify individual scene components & structures of multiple elements
This capability will be critical in futureUGV systems to fully comprehend the local environment – for mobility & behaviors.
Input scene Classification output of 3D points (red: surfaces, green: vegetation, blue:
elongated structures).
RoboticsCollaborative Technology Alliance
Focused Initiatives
RoboticsCollaborative Technology Alliance
Focused Initiatives
Mid-range Sensing Mid-range Sensing
•Developed techniques to estimate the range of structures out to a kilometer from robot
•Permits “coarse” estimation of the environment for planning.
• “Visual odometry” process aids vehicle navigation, especially under GPS denied conditions.
vehicle path
reconstructed 3D features
vehicle path
reconstructed 3D features
vehicle path
reconstructed 3D features
Using very accurate determination of vehicle position and orientation, locations of objects in the scene can be determined by triangulation applied to video images.
RoboticsCollaborative Technology Alliance
Focused Initiatives
RoboticsCollaborative Technology Alliance
Focused Initiatives
Tactical Behaviors Tactical Behaviors
Implemented the 4D/RCS architectural framework to provide capabilities to:
•Support deliberate and reactive tactical behaviors to include tactical skills, individual tasks and collective tasks;
•Support implementation of autonomous command and control of behaviors associated with execution of “move”, “look”, “shoot”, and “communicate” tasks;
•Span multiple levels of control from servo up to platoon including ensemble (unmanned air and ground vehicles and unattended ground sensors) unmanned battle teams;
•Support variable levels of operator control from
supervisory to fully manual;
•Adaptive behaviors for dynamic environments
SPWMBG
SPWMBG
SPWMBG
SPWMBG
SPWMBG SPWMBG
SPWMBG
SPWMBGSPWMBG
SPWMBGSPWMBG SPWMBG
SPWMBGSPWMBGSPWMBG SPWMBGSPWMBG SPWMBGSPWMBGSPWMBG
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
To TARDEC’s Vetronics Technology Integration program
To TARDEC’s Vetronics Technology Integration program
•All hardware and software
•Sensor processing algorithms,
•Vehicle planners
•RCTA's 4D Real Time Control System (4D-RCS) intelligent system control architecture
RCTA technology transferred to the 18-ton Stryker
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
To Future Combat Systems Autonomous Navigation System program
To Future Combat Systems Autonomous Navigation System program
•Field tested laser scanner hardware
•Laser processing algorithms for obstacle detection and classification
•Engineering visualization tools for laser and vehicle planner development
•Operating robotic testbed platforms (with interfaces to navigation sensors) fully capable of data collection and archiving in realistic tactical environments
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
RoboticsCollaborative Technology Alliance
Transitions and Tech Transfers
To Natick Soldier Center’s Future Force Warrior program To Natick Soldier Center’s Future Force Warrior program
•Operational Command Language for specifying mission tasks
•Soldier’s Decision Support System for command and control of robotic vehicles
•Tactical behaviors for robotic platforms performing recon, surveillance and target acquisition (RSTA) missions
•Command and Control/Soldier Machine Interface API
•Spoken Language Interface for hands-free control of robotic assets
•Single-screen OCU for both autonomous and direct Soldier control of robotic assets
•Autonomous navigation system architecture
•Collaborative air-ground behaviors
•Experimentation with MULE class platforms
RoboticsCollaborative Technology Alliance
Developing Advanced Capabilities to Meet
the Army’s Vision
RoboticsCollaborative Technology Alliance
Developing Advanced Capabilities to Meet
the Army’s Vision
From teleoperation of counter-mine vehicles
in Iraq todayTo integration of
unmanned systems for dismounted operations
for the Future Force
To fully autonomous mobility for manned
and unmanned vehicles in the Army’s Future
Combat System