DARPA Robotics Challenge Proposers’ Daycga/humanoids-ugrad-12/drc.pdf · 2 . DARPA Robotics...

DARPA Robotics

Challenge

Proposers’ Day April 16, 2012

2

DARPA Robotics Challenge Proposers’ Day

The Proposers’ Day will

begin at 12:10 PM EDT

April 16, 2012

3


Dr. Gill Pratt Program Manager, DARPA

DARPA Robotics Challenge Proposers’ Day Agenda

4

12:00PM-12:10PM Set up/Test Connection

12:10PM-12:15PM Introductory Remarks, OSTP Dr. John Holdren, Assistant to the President for Science and Technology and Director of Science and Technology Policy (OSTP)

12:15PM-12:25PM Introductory Remarks, DARPA

Dr. Kaigham J. Gabriel, Acting Director, DARPA

12:25PM-12:40PM DARPA/TTO Overview Mr. Paul Eremenko, Acting Director/Deputy Director, DARPA TTO

12:40PM-1:40PM Robotics Challenge Overview Dr. Gill Pratt, Program Manager, DARPA

1:40PM-1:55PM Contracts Management Mr. Chris Glista, Contracting Officer, DARPA

1:55PM-2:10PM GFE Hardware Dr. Robert Playter, Boston Dynamics, Inc.

2:10PM-2:25PM GFE Software Mr. Nate Koenig, Open Source Robotics Foundation, Inc.

2:25PM-2:45PM Break

2:45PM-3:15PM Q/A Response

3:15PM Adjourn

5

Introductory Remarks, OSTP

Dr. John Holdren Assistant to the President for Science and Technology and Director, Office of Science and Technology Policy (OSTP)

Dr. John Holdren VIDEO

4/16/2012 6 Further dissemination only as directed by DARPA Public Release Center or higher DoD authority.

7

Introductory Remarks, DARPA

Dr. Kaigham J. Gabriel Acting Director, DARPA

4/16/2012 8 Further dissemination only as directed by DARPA Public Release Center or higher DoD authority.

Dr. Kaigham J. Gabriel VIDEO

DARPA/TTO Overview

Mr. Paul Eremenko, Acting Director Tactical Technology Office

Briefing prepared for Robotics Challenge Proposers’ Day

April 16, 2012 Photo: DARPA

4/16/2012 9 Distribution Statement “A” (Approved for Public Release, Distribution Unlimited).

1 9 5 7

Sputnik

34th President of the United States 1953-1961. Coined the term “military-industrial complex” and warned against its unwarranted influence. Created DARPA in response to Sputnik.

First orbiting satellite. The satellite was not a threat, but the level of technology indicated that the Soviet Union possessed superior capability for intercontinental reconnaissance and bombing.

Dwight D. Eisenhower

1 9 5 8

DARPA


The Defense Advanced Research Projects Agency (DARPA) was established in 1958 to prevent strategic surprise from negatively impacting U.S. national security and create strategic surprise for U.S. adversaries by maintaining the technological superiority of the U.S. military. To fulfill its mission, the Agency relies on diverse performers to apply multi-disciplinary approaches to both advance knowledge through basic research and create innovative technologies that address current practical problems through applied research. As the DoD’s primary innovation engine, DARPA undertakes projects that are finite in duration but that create lasting revolutionary change.

Mission


► Create and prevent strategic surprise

► Lean, agile risk-taking organization

► Idea-driven and outcome-oriented

► Demonstrate solution concepts

Idea leadership culture


“Impossible” ideas become real life at DARPA


1 8 6 5 1 9 6 7 1 9 5 7

DARPA: F1 engine Saturn V

Photo: http://moonbooks.net

Books

Spaceflight


1 9 6 6 1 9 8 6 1 9 8 0 ’ s

Photo: http://www.globalnerdy.com Photo: http://www.wired.com

Television DARPA: research Cellular phone

Wireless communications


Original ARPAnet for remotely sharing large volumes of data.

1 9 6 9 1 9 6 9 Fictional access news and personal communication on tablet devices.

2 0 1 0

News and personal communication available on tablet devices.

Photo: http://www.mercurybrief.com/ 2010/01/kubrick-slates-tablets-2001-film/

Photo: http://www.surfmobee.com

The internet


DARPA sponsored flight technology demonstrator to improve survivability of a single aircraft against a modern IADS

Combined with precise weapons against C2 targets to improve survivability for all forces in

Desert Storm

1 9 9 1 1 9 7 8 1 9 7 6

“Hopeless diamond” concept rejected by U.S. Air Force

“Stealth”


1 9 9 0 ’ s 1 9 6 3 DARPA Transit 2A satellite

pioneered doppler navigation for specific military missions.

Became global precision navigation technology for military and civilian use.

Continued military and commercial refinement for

tactical applications

2 0 1 0

Global Positioning System


1 9 9 7 - 2 0 0 9 1 9 8 6

DARPA flight technology demonstrator to prove the feasibility of the technology; shelved because it did

not fit into prevailing air operations thinking.

Rapidly adopted and matured to counter new threats more effectively than existing

platforms and doctrine could.

Photo: General Atomics

Photo: General Atomics

Surveillance UAV’s


2 0 1 1 1 9 2 7

DARPA: Autonomous Robotics Manipulation

Image: http://www.jeffbots.com/maria.html

2 0 X X

Movies

Automation


1 9 6 2 2 0 1 0 2 0 X X

Books Research

Image: http://www.tomswift.info

Multi-mode mobility


“X”-Planes

RF Stealth The Internet

GPS

Space Launch and Navigation

Synthetic Aperture Radar

Autonomous Systems

Continuous accomplishment


• Comms & Networks

• Global Tactical ISR

• Energy • Hybrid

Warfare • • Extreme

Environments

• Space Systems & Space Awareness

• Hypersonics

• Space Access

• Aero- & Hydrodynamic Systems

• Ground & Soldier

Systems • Design Synthesis • Manufacturing • Qualification/V&V • Autonomy

• Global ISR

• Cyber

• Social Networks

• Computational Social Science

• Language Transparency

• Edge Finding

• Training/ Education

• Basic Science Core

• Devices

• Integration

• Power

• Architectures

• Application

• Physical Sciences

• Materials

• Mathematics

• Training & Human Effectiveness

• Biological

Warfare Defense

• Biology

DSO MTO I2O TTO STO AEO

• Agile Programs with Frequent Development Cycles

• Conduct Systematic Rigorous

• Assessments

• Explore New Contracting Approaches

• Develop Strong Relationships

Defense Sciences Office

Microsystems Technology Office

Information Innovation Office

Tactical Technology Office

Adaptive Execution Office

Strategic Technology Office

Organization


Flat organization Little hierarchy to ensure the free and rapid flow of information, ideas, and decisions.

Outstanding program managers DARPA hires creative, independent people with big ideas and empowers them.

Project-based assignments Projects organized around a challenge model and typically last three to five years; longer if necessary to facilitate transition.

No DARPA labs Majority of the research is sponsored in industry and universities with a small amount in government labs.

Flexible outsourcing of staff and performers Great talents and ideas from industry, universities, and government labs with technical, contracting, and administrative services from other commercial and government agencies.

Professional reach


www.darpa.mil


DARPA Robotics Challenge

Proposer’s Day

April 16th, 2012

Dr. Gill A. Pratt, Program Manager

26 Distribution Statement "A" (Approved for Public Release, Distribution Unlimited)

Recent Disasters

27 Distribution Statement "A" (Approved for Public Release, Distribution Unlimited) Fukushima - 2011

Katrina - 2005 Deepwater Horizon

2010

Inspiration and Goals


1. Target disaster response in dangerous

environments, and important DoD capability for

HADR (Humanitarian Assistance and Disaster

Relief) missions

2. Advance supervised autonomy, mobility,

manipulation, and energetic efficiency.

3. Catalyze the robotics industry by developing a

validated, real-time, operator-interactive

simulator.

4. Welcome a wide range of international

contributors including traditional and non-

traditional DARPA performers from a variety of

fields.

Man-Computer Symbiosis

J. C. R. Licklider (head of DARPA IPTO 1962) IRE Transactions on Human Factors in Electronics,

volume HFE-1, pages 4-11, March 1960

“… close study of the disaster’s first 24 hours, before the cascade of failures carried reactor 1 beyond any hope of salvation, reveals clear inflection points where minor differences would have prevented events from spiraling out of control.” IEEE Spectrum, November 2011 pg. 36.

Fukushima - 2011

1. Environment, even degraded, has been engineered for humans

2. No shortage of human tools, from hand tools to vehicles

3. Human-like robot capabilities are easier for domain experts to understand and untrained operators to control

Why emphasize human-like robot capabilities?


30

Supervisory Autonomy and Predictive Models

Environment Human

Environment Tele-op Robot

Human

Environment Tele-op Robot

Environment Human Predictive Model

Environment Sprv. Auto Robot

Sprv. Auto Robot

Environment Human

DARPA Challenge

High Fidelity Interface

Low Fidelity Interface

Variable Fidelity Interface

Predictive Model

Human Predictive Model


Decouple

Decouple

Distribution Statement "A" (Approved for Public Release, Distribution Unlimited)

31

Questions this Program will Address


Sprv. Auto Robot

1. What kind of I/O

2. What kind of messages

3. What autonomy, mobility, manipulation, energetics

4. Tasks and Metrics


32

Example Disaster Challenge Scenario

Auto

nom

y - P

erce

ptio

n

Auto

nom

y –

Deci

sion-

mak

ing

Mou

nted

Mob

ility

Dism

ount

ed M

obili

ty

Dext

erity

Stre

ngth

Endu

ranc

e

1. Drive utility vehicle to site X X X X

2. Travel dismounted across rubble X X X

3. Remove debris blocking entryway X X X X X

4. Open door, enter building X X X X

5. Climb industrial ladder, traverse industrial walkway X X X

6. Use tool to break through concrete panel X X X X X

7. Locate and close valve near leaking pipe X X X X X X

8. Replace component X X X

Task

s

Capabilities


33

Planned Tasks – subject to change

Task 1 - Drive a utility vehicle to the site The robot must demonstrate mounted mobility by ingress to the vehicle, driving it on a road, and egress from the vehicle. The robot must also demonstrate manipulation by operating the controls, including steering, throttle, brakes, and ignition. The robot must steer, accelerate, and brake. Task 2 -Travel dismounted across rubble The robot must demonstrate dismounted mobility by crossing terrain ranging from smooth and level, to rough and sloped, with some loose soil and rocks. This terrain will be easily traversable by a human. Task 3 - Remove debris blocking an entryway The robot must demonstrate the dexterity and strength to move an object blocking an entryway. The object will have size, weight, and other properties to be movable either by a person or by the GFE Platform. The object mass is expected not to exceed 5 kg. Task 4 - Open a door and enter a building The robot must demonstrate the dexterity to operate a door handle and the strength to push the door open.


34

Planned Tasks (cont.) – subject to change

Task 5 - Climb an industrial ladder and traverse an industrial walkway The robot must demonstrate dismounted mobility to traverse an industrial elevated walkway. It is expected that the walkway (or catwalk) will have a grated surface and handrails. Task 6 - Use a tool to break through a concrete panel The robot must demonstrate using a power tool to perform forceful manipulation. The power tool is expected to be an air or electric impact hammer and chisel, or an electric reciprocating saw. Task 7 - Locate and close valve near leaking pipe The robot must demonstrate the perception ability to find a leaking pipe and a nearby valve, the dismounted mobility to approach the valve, and the manipulation ability to close the valve Task 8 - Replace cooling pump The robot must demonstrate the perception ability to locate the pump, the manipulation ability to loosen one or more fasteners, and the bi-manual manipulation ability to extract and replace the pump.


Distribution Statement "A" (Approved for Public Release, Distribution Unlimited) 35

Participation Options

Track A - Funded Teams design and build platform systems to participate in Disaster Response Challenge. Track B –– Teams develop algorithms using DARPA provided Simulator. Disaster Virtual Challenge conducted to identify which teams will be provided GFE Robot and additional funding to participate in Disaster Response Challenge Track C – Unfunded teams use Simulator to and develop algorithms and participate in Disaster Virtual Challenge and will compete along with Track B teams for GFE Robot and funding award Track D – Unfunded teams develop platforms to compete in Disaster Response Challenges

Planned Schedule and Funding


37

Planned Program Funding

The program has planned the following funding for teams. This plan is subject to change depending on the number of qualified teams and available resources. Phase 1 Track A: Up to $3M for each team (up to five (5) teams for fifteen (15) months). Track B: first nine (9) months through the Virtual Disaster Response Challenge: Up to $375k for each team, with up to twelve (12) teams. Track B/Track C: last six (6) months after the Virtual Disaster Response Challenge: Up to $750k additional for each team, with up to six (6) teams. Phase 2 Up to $1M for up to eight (8) of the top performing teams from the first physical disaster challenge. Prize Anticipated to be $2M to a single team.


Planned Key Dates

38

• May 31, 2012 – BAA Closing

• October 2012 – Anticipated Contract Awards/Program Kickoff

• June 2013 – Virtual Disaster Response Challenge

• December 2013 – Disaster Response Challenge #1

• December 2014 – Disaster Response Challenge #2


www.darpa.mil


BAA Process

Mr. Chris Glista

DARPA Contracts Management Office


April 16, 2012

40 Distribution Statement A

• BAA and any amendments posted in FEDBIZOPPS

• BAA covers all info needed to propose

• TIME PERIOD – BAA is open until May 31, 2012

• Instructions are detailed in the BAA (Follow closely)

• Following the proposal instructions assists the evaluation team to clearly understand what is being proposed and supports a timely negotiation.

• ALL questions to [email protected]

BAA Specifics

4/16/12 41 Distribution Statement A

• Federal Acquisition Regulation (FAR) Part 35

http://farsite.hill.af.mil

• “A general announcement of an agency’s research interest including criteria for selecting proposals of all offerors capable of satisfying the Government’s needs.”

• For the acquisition of basic and applied research not related to the development of a specific system or hardware procurement.

BAA Background


http://farsite.hill.af.mil/

• Used when proposals with varying technical and scientific approaches can be expected

• Proposals need not be evaluated against one another since they are not submitted in accordance with a common work statement.

• Primary basis for selecting proposals shall be technical, importance to agency programs, and fund availability.

• Cost realism shall also be considered

BAA Background (cont.)


• Contractors SHALL NOT subcontract technical or scientific work without the contracting officer’s knowledge

BAA Background (cont.)


• FAR PART 15 “Contracting By Negotiation” does not apply in full

• No “competitive range” determination

• No mandatory “discussions”

• No opportunity for proposal revision

• “Cost and Pricing Data” (certification required for all FAR-based contract proposals greater than $700K)

• Subcontract consent and key personnel clauses to be included in awards

IMPLICATIONS


• ALL questions to [email protected]

• After Receipt of Proposals – Government (PM/PCO) may communicate with proposers to understand the meaning of some aspect of the proposal that is not clear or to obtain confirmation or substantiation of a proposed approach, solution, or cost estimate

• Informal feedback may be provided once selection(s) are made – no “ranking” provided

Communications


mailto:[email protected]

• All interested/qualified sources may respond

• International participants/resources may participate to the extent authorized by applicable Security Regulations, Export Laws, etc.

• Small Business Participation Encouraged, Teaming Not Required

Applicant Eligibility


• Government agencies/labs, FFRDCs cannot propose to this BAA in any capacity, UNLESS they can clearly demonstrate the work is not otherwise available from the private sector AND they also provide written documentation citing the specific statutory authority (as well as, where relevant, contractual authority) establishing their eligibility to propose to government solicitations

Applicant Eligibility (cont.)


• All facts relative to a potential conflict of interest must be disclosed.

• Examples of conflicts listed at FAR Part 9.505

• Provided Systems Engineering and Technical Direction

• Contractors MAY NOT simultaneously provide SETA support and be a technical performer at DARPA without waiver approved by DARPA Director

Organizational Conflict of Interest


Award Instrument

• Contract, Grant, Cooperative Agreement, Other Transaction

• http://farsite.hill.af.mil/ (FAR Contracts)

• http://www.dtic.mil/whs/directives/corres/html/321006r.htm (Grants, Cooperative Agreements)

• www.acq.osd.mil/dpap/Docs/otguide.doc (Other Transactions)

• In all cases the Contracting Officer shall have sole discretion to select award type instrument

Award Type

4/16/12 50 Distribution Statement

http://farsite.hill.af.mil/

http://www.dtic.mil/whs/directives/corres/html/321006r.htm

http://www.acq.osd.mil/dpap/Docs/otguide.doc

• Assert rights to all technical data & computer software generated, developed, and/or delivered to which the Government will receive less than Unlimited Rights

• Assertions apply to Prime and Subs

• Justify “Basis of Assertion”

• Use defined “Basis of Assertion” and “Rights Category”

• Assessed during evaluation under the “Potential to Accomplish Technology Transition” Evaluation Factor

Rights to Technical Data and Computer Software


Atlas Robot Overview for DARPA Robotics Challenge

Robert Playter Marc Raibert April 16, 2012

Atlas Overview

• Near copy of

PETMAN/Atlas • 50th percentile male

anthropometry • 27 actuated, hydraulic

DOF’s • 2 on-board computers • Power tether • Sensing head • Modular wrist • Protective shells

Current Performance

PETMAN Video

Robot Degrees of Freedom (DOF)

• 27 Active DOFs – NECK: 2 DOF – BACK: 3 DOF – SHOULDER: 3 DOF – ELBOW: 1 DOF – WRIST: 2 DOF – HIP: 3 DOF – KNEE: 1 DOF – ANKLE: 1 DOF

• Hands attach through modular wrist – Mechanical interface – Electrical & data interface

56

Actuator Design Targets

Joint DoF Derived RoM in degrees

Adduction/Abduction (° rx) Flexion/Extension (° ry) Internal/Medial Rotation (° rz) min max range min max range min max range

hip 3 -30 30 60 -100 20 120 -10 70 80 knee 1 0 135 135 ankle 2 -40 40 80 mid foot n/a back 3 -40 40 80 -35 25 60 -45 45 90 neck 3 -45 45 90 -90 90 180 shoulder 3 0 115 115 -90 50 140 -40 100 140 elbow 1 -135 0 135 wrist 3 -30 60 90 -85 85 170

Joint DoF Derived Strengths in Newton-Meters

Adduction/Abduction (rx) Flexion/Extension (ry) Internal/Medial Rotation (rz) Strength Strength Strength

hip 3 180 260 110 knee 1 220 ankle 2 220 mid foot 0 Passive Flexure back 3 100 180 110 neck 3 25 54 shoulder 3 110 110 80 elbow 1 100 wrist 3 30 15

57

Robot Weight Estimates

• Weight estimate for current design: 178 lbs

COMPONENT MASS [KG]

MASS [LBS]

Pelvis & Leg 30.7 67.7

Torso & Back 9.8 21.5

Arm & Hand 10.2 22.6

Head & Neck 1.8 4.0

Hoses & Cables 3.4 7.6

Skin 17.0 37.5

Misc. Hardware 5.5 12.1

Fluid (Oil/Water) 2.5 5.6

ROBOT TOTAL 81 178

Power Plant 22.0 48.5

TOTAL 103 227 12%13%

2%

4%

21%

7%3%

38%

Pelvis & LegTorso & BackArm & HandHead & NeckHoses & CablesSkinMisc. HardwareFluid (Oil/Water)

58

Power Pack

• Tether – Electric power – Cooling water

Power Tether

Laboratory Electrical

Power

Cooling Water

Robot Computing Architecture

Robot Computer Actuators (27) Joint Sensors Hands

GigE

CANbus

GigE

User Computer

Sensors - IMU - LIDAR - Stereo Cameras

CANbus

WIFI

Mr. Nate Koenig Open Source Robotics Foundation, Inc.

Simulation for robots

Focus on accurate physical simulation Easy transition to and from simulation Remove hardware issues and resource constraints

Support common robot control software Custom client code ROS interface Player interface

Support sharing of resources New sensors, actuators, models, and environments

Additional Benefits

No real-time constraints Simulate faster than real time

Regression testing Use simulation for automated tests

Universal test environment Create benchmarks Run a competition

Flexibility Handles a wide range of environments and tasks Thin programmatic layer to Gazebo functionality

Architecture

Physics Visualization Interfaces

Rigid Body Dynamics OpenGL Plugins and IPC ODE Bullet

OGRE Google Protobuf Boost

Robot Models

Simple platforms Built-in shapes Mesh skinning

Realistic physical properties Meshes as collision objects Mass and inertia properties Surface friction 6 joint types

Full sensor suite Laser range finders Mono/Stereo cameras Kinect Contact Joint force/torques

Environments Simple

Focused scenario Manipulation Perception

Aerial robots Outdoor mobile and legged robots

Outdoor

Indoor Path planning Mobile manipulation Clone real environment

Show video with next slide Presenter will talk to it

Gazebo Demo Part 1: Canyon fly-through

Custom terrain generated from a greyscale image Animated quadrotor

Thanks to Johannes Meyer and Stefan Kohlbrecher

Part 2: Pioneer2dx and office environment Player interface used to drive the Pioneer2dx Laser range finder sensor visualization

Part 3: PR2 and YouBot PR2 object manipulation using ROS Interactive markers

Part 4: Character animation Experimental animation of characters using skeletons Thanks to Mihai Dohla

Tools

Command line tools System inspection Insert and remove models

Graphical tools Model placement Joint and mass visualizations Sensor visualizations

Resources

Robot models Distributed in Gazebo Work in progress

Online model database Graphical model builder

Environments Distributed in Gazebo Google's 3D warehouse Google Sketchup or Blender

Help http://gazebosim.org

Roadmap

1.2

Vehicles Tools Outdoor environments Cluster support

1.4

Bullet integration Multi-floor structures Sonar, force-torque, pressure sensors

1.6

Customize GUI API Large environments Vehicle suspension models

2.0

Friction and noise models Hardware in the loop Deformable objects

Community

Support and Contributing http://gazebosim.org/support.html

Wiki http://gazebosim.org/wiki

Kforge project https://kforge.ros.org/projects/gazebo/

Thank you

Physics Lead John Hsu

Co-Founder: Retired Andrew Howard

Core Contributor Mihai Dolha

Contributors Christian Gagneraud Brian Gerkey Stefan Kohlbrecher Johannes Meyer Alexis Maldonado Jordi Polo Abbas Sadat Richard Vaughan

78


We are on a temporary break

and will be back shortly

79


Q&A Response

80


Adjourn

DARPA Robotics Challenge Proposers’ Daycga/humanoids-ugrad-12/drc.pdf · 2 . DARPA Robotics...

Documents

Transcript of DARPA Robotics Challenge Proposers’ Daycga/humanoids-ugrad-12/drc.pdf · 2 . DARPA Robotics...