2012-12-17

Open Source Robotics Foundation, Inc

12.8: news.adds/robotics/Open Source Robotics Foundation, Inc
2012.4.10, 30: DRC (Robotics Challenge) Broad Agency Announcement:
GFE Simulator is expected to be provided by the
Open Source Robotics Foundation, Inc.,
and will initially be based on the ROS Gazebo simulator.
Expectations for the GFE Simulator include the following:
• Models the three-dimensional environment;
• Allows import of robot's
kinematic, dynamic, and sensor models;
• Allows robot's co.workers to
send commands over a network
(identical to those sent to a physical robot)
and receive data from the simulated robot
(similar to that received from a physical robot);
• Uses physics-based models of
inertia, actuation, contact, and environment dynamics
to simulate the robot’s motion;
• Runs in real-time on the “cloud,”
likely on Graphics Processing Units (GPUs);
• usable from the cloud
by at least 100 concurrent teams
The GFE Simulator supplier
will manage an open-source effort
where the simulator, robot models,
and environment models
are developed and improved by the supplier
as well as by contributors throughout the world.
2012.4.12: OSRF @ fbo.gov:
Sole Source Intent Notice
Open-Source Robot SImulation Software
for DARPA Robotics Challenge
Solicitation Number: DARPA-SN-12-34

The Defense Advanced Research Projects Agency,
Tactical Technology Office (TTO),
intends to award a sole source contract to the
Open Source Robotics Foundation, Inc.,
of Menlo Park, California.
For the contract, it will develop an open-source
robot simulation software system
for use by the DARPA Robotics Challenge program.
The effort will develop validated models of
robots and field environments  .
The effort will make the simulation software
available on an open-source basis,
and will host the simulation so that
participants in the DARPA Robotics Challenge program
can access it freely.
To search for innovative approaches to provide this service,
DARPA performed an informal market survey.
Of the few existing candidate open-source robot simulators,
the Open Source Robotics Foundation was deemed to be
the sole viable supplier for providing the
necessary open-source simulation software
within the specified timeframe.
The Open Source Robotics Foundation simulator, called Gazebo,
simulates many robots and sensors,
supports the ROS and Player robot control middleware,
and maintains tutorials, documentation
and an active user base.
The technology underlying Gazebo
supports many use cases
including remote operation
through a client-server architecture,
customization via a flexible plugin interface,
accurate physics simulation,
and rapid setup through intuitive
graphical and programmatic interfaces.
The proposed contract action is for
supplies or services for which the Government
intends to solicit and negotiate with
only one source under authority of FAR 6.302-1
"Only one responsible source
and no other supplies or services
will satisfy agency requirements."
As the open-source robot simulation leader,
the Open Source Robotics Foundation possesses
unique knowledge and capabilities required to
carry out the required research effort.
No other source would be capable of
satisfying the requirements
for an affordable end-to-end solution
necessary to meet the Government's needs.
2012.4.16: OSRF @ willowgarage:
The history of open source is replete with
significant moments in time,
and today Willow Garage would like to humbly submit
our own milestone -- the announcement of
the Open Source Robotics Foundation.
As Willow Garage has worked to grow and shepherd ROS
within the robotics world
our hope was that ROS would one day stand on its own.
With the announcement today of the OSR Foundation,
that day is finally here.
The mission of the OSR Foundation is "to support
the development, distribution, and adoption
of open source software for use in robotics
research, education, and product development."
The first initiative of the OSRF will be participation in
the DARPA Robotics Challenge, announced recently.
DARPA today sponsored a Proposer's Day Workshop
where more information about the Robotics Challenge
is available via Webcast.
During the Webcast, Nate Koenig from Willow Garage
gave a brief talk on the current and future state
of the open source Gazebo robot simulator,
which will be extended by the OSR Foundation
to support the DARPA Robot Challenge.
2012.4.17: Gazebo @ spectrum.ieee.org:
You're likely already familiar with the
Robot Operating System, or ROS,
in relation to Willow Garage's PR2 robots.
A few years ago, Willow Garage
integrated ROS and the PR2 into Gazebo,
a multi-robot simulator project
started at the University of Southern California
by Andrew Howard and Nate Koenig.
Willow Garage now provides financial support
for the development of Gazebo.
. ROS is now mature enough to go off and fend for itself,
and the Open Source Robotics Foundation
is the shiny new embodiment of that confidence.
DARPA isn't awarding the simulator contract to
Willow Garage or Gazebo's org,
but to a new foundation dedicated to the
full lifecycle of Robotics programming .
. here's what DARPA is expecting:
    The Open Source Robotics Foundation
    will develop an open-source
robot simulation software system
    for use by the DARPA Robotics Challenge program.
    The effort will develop validated models of robots
    (kinematics, dynamics, and sensors)
    and field environments (three-dimensional
    surfaces, solids, and material properties).
    The effort will develop physics-based models of
    inertia, actuation, contact, and environment dynamics
    to simulate the robot's motion.
    . available on an open-source basis,
    and hosted so that participants in the
    DARPA Robotics Challenge program
    can access it freely.
Open Source Robotics Foundation`sponsored projects:
. Gazebo is a 3D multi-robot simulator with dynamics.
. ROS (Robot Operating System) provides
hardware abstraction, device drivers, libraries, visualizers,
message-passing, package management, and more.
. Paradise Studios SkyX(open version) provides a
photorealistic, simple and fast sky rendering plugin
for the Ogre3D graphics engine,
compatible with Gazebo .
. Other open-source activities
that are related to the mission of OSRF
include the Arduino platform and the Make database.
. sponsors include:
# Bosch's Research and Technology Center (RTC)
has the distinction of being the only
non-university recipient of a Willow Garage PR2
as part of the PR2 Beta Program
[for hardware-testing with the ROS .]
# DARPA (Defense Advanced Research Projects Agency)
is funding the Gazebo robot simulator,
which will be used extensively in the
DARPA Robotics Challenge.
# Rethink Robotics's Baxter robot uses ROS .
# Sandia National Laboratories is
collaborating with OSRF to develop
low-cost, highly dexterous robot hands.
# Willow Garage is a Founding Contributor to OSRF.
Without Willow Garage's financial,
organizational, and moral support,
OSRF could never have become
the company that we are today.
# Yujin Robot is collaborating with OSRF,
to develop tools for authoring, and management of
multi-robot systems in production environments .
2012.7: OSRF will concentrate on Gazebo:
. we're planning to put a lot of work into
the Gazebo simulator and ROS,
significantly improving the capability
and availability of robot simulation .
. we can have greater impact by
taking a "deep dive" in one area
rather than taking shallow responsibility for everything.
Over time, we'll contribute to a broader variety of projects,
mostly drawn from the ROS ecosystem.
. a ROS certification program may be
most valuable in a domain-specific manner.
eg, imagine a way of certifying a
ROS-controlled industrial robot arm
as being compatible with an accepted
ROS interface standard
(perhaps crafted by the forthcoming
ROS Industrial Consortium .
Similarly, imagine a TurtleBot being certified
as implementing REP 119 .
. compatibility should be the focus;
certifying functionality is an entirely different topic .
. the ROS Platform development
continues to be managed by Willow Garage,
but responsibility for release management
will transition to OSRF eventually .
DRC/Roadmap > Tutorials > Main Page > Overview:
Gazebo is a multi-robot simulator for outdoor environments.
Like Stage (part of the Player project),
it is capable of simulating a population of
robots, sensors and objects,
but does so in a three-dimensional world.
It generates both realistic sensor feedback
and physically plausible interactions between objects
(it includes an accurate simulation of rigid-body physics).
Why use Gazebo?
Gazebo was originally designed to develope
algorithms for robotic platforms.
By realistically simulating robots and environments
code designed to operate a physical robot
can be executed on an artificial version.
This helps to avoid common problems
associated with hardware
such as short battery life, hardware failures,
and unexpected and dangerous behaviors.
It is also much faster to
spin up a simulation engine
than continually run code on robot,
especially when the simulation engine
can run faster than real-time.
Over the years Gazebo has also
been used for regression testing.
Scenarios designed to test algorithm functionality
have been established and passed through test rigs.
These tests can be run continually
to maintain code quality and functionality.
Numerous researchers have also used Gazebo
to develop and run experiments
solely in a simulated environment.
Controlled experimental setups
can easily be created in which subjects can
interact with robots in a realistic manner.
Additionally,
Gazebo has been used to compare algorithms
for such things as navigation
and grasping in a controlled environment.
Future:
Gazebo is under active development at Willow Garage.
We are continually fixing bugs,
and adding new features.
If you have feature requests, need help,
or have bugs to report
please refer to the support page .
gazebosim.org`features:
Gazebo is a feature rich application that is
under constant development from a large user community.
The follow lists a few of the primary features offered by Gazebo.
Dynamics Simulation
    Access multiple physics engines including ODE and Bullet.
    Direct control over physic engine parameters
    including accuracy, performance, and dynamic properties
    for each simulated model.
Advanced 3D Graphics
    Utlizing OGRE, Gazebo provides realistic
rendering of environments.
    State-of-the-art GPU shaders
generate correct lighting and shadows
    for improved realism.
Sensors
    Generate sensor information from
    laser range finders, 2D cameras, Kinect style sensors,
    contact sensors, and RFID sensors.
Robot Models
    Many robots are provided including
    PR2, Pioneer2 DX, iRobot Create, TurtleBot,
    generic robot arms and grippers.
Environments
    Access to many objects from simple shapes to terrain.
Programmatic Interfaces
    Support for ROS and Player.
    API for custom interfaces.
Plugins
    Develop custom plugins for robot model control,
    interacting with world components
    Provides direct control to all aspects of
the simulation engine including
    the phyics engine, graphics libraries,
and sensor generation
TCP/IP Communication
    Run Gazebo on remote servers
    Interface to Gazebo through
socket-based message passing
    using Google Protobufs.
Powerful Graphical Interface
    A lightweight QT based graphical interface
    provides users with direct control over
    many simulation parameters.
    View and navigate through a running simulation.
Collada Import
    Import meshes from many sources
    using Gazebo's built in Collada reader.
Active User Community
    Research institutes around the world
utilize and contribute to Gazebo.
    Community supported help
through a mailing list, and wiki.
Person simulation
    Replay human motion capture data
in a running simulation.
install on Ubuntu Linux 12.04 (precise):
Setup your computer to accept software from
packages.osrfoundation.org.
$ sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu
precise main" > /etc/apt/sources.list.d/ros-latest.list'
$ sudo sh -c 'echo "deb http://packages.osrfoundation.org/gazebo/ubuntu
precise main" > /etc/apt/sources.list.d/gazebo-latest.list'

Retrieve and install the keys:
$ wget http://packages.ros.org/ros.key -O - | sudo apt-key add -
$ wget http://packages.osrfoundation.org/gazebo.key -O - | sudo apt-key add -

Update apt-get and install Gazebo:
sudo apt-get update
sudo apt-get install gazebo

source: drc/ubuntu/pool/main/g/gazebo .

seems to speak c++/boost:
Create a model plugin that subscribes to a ROS topic:
#include
#include
...
namespace gazebo {     
  class ROSModelPlugin : public ModelPlugin   { ... .
-- LOOK C++ !!!
. this may be why DARPA said
Gazebo is only a tentative choice;
they might have plans to translate all this work
over to some better programming language
supporting the new systems in the pipeline
that feature improved massive concurrency
based on very energy-efficient chips ... .

darpa's new project is UHPC 2010:
(Ubiquitous High Performance Computing)
The UHPC program seeks to develop
the architectures and technologies
that will provide the underpinnings,
framework and approaches
for improving power consumption, cyber resiliency,
and programmer productivity,
while delivering a thousand-fold increase
in processing capabilities.
2012.6.25: UHPC details:
. DARPA, in 2010, selected four "performers"
to develop prototype systems for its UHPC program:
Intel, NVIDIA, MIT, and Sandia National Laboratory.
. NVIDIA is teaming with Cray [designer of Chapel.lang],
Oak Ridge National Laboratory and others
to design its ExtremeScale prototype.
NVIDIA was not previously seen as
the cutting-edge of supercomputing,
[but GPU's are apparently the future
of massive concurrency ].
. Intel's 2010 revival of Larrabee,
as the Many Integrated Core (MIC) processor,
may be of use in its UHPC designs .
. Intel's prior concurrency tools include Ct language,
Threading Building Blocks, and Parallel Studio .
The first UHPC prototype systems
are slated to be completed in 2018.
. single-cabinet UHPC systems will need to deliver
a petaflop of High Performance Linpack (HPL)
and achieve an energy efficiency of at least
50 gigaflops/watt (100 times more efficient
than today's supercomputers.)
. the programmer should be able to implicitly
implement parallelism (without using MPI
or any other communication-based mechanism).
In addition, the operating system for these machines
has to be "self-aware,"
such that it can dynamically manage performance,
dependability and system resources.

No comments:

Post a Comment