IEEE Robotics & Automation Magazine - March 2014 - 20

does not directly fund projects, but
does have significant influence, and in
2012 found a way to support ROS.
Brian Gerkey, who had led the software
efforts at Willow Garage, founded the
Open Source Robotics Foundation
with initial support from Willow, but
grew and sustained it with a large grant
from DARPA for the simulation part of
the DARPA robotics challenge.
Gary Bradski, the creator of
OpenCV while at Intel, had convinced
Willow Garage management to sponsor the OpenCV, since it had been stagnant without significant funding since
2008. After four years of renewed funding from Willow Garage, OpenCV was
going strong and garnering interest
from other industry players. When he
left Willow Garage to join IPI, Gary
also founded the OpenCV Foundation
to ensure continued support for
OpenCV beyond Willow Garage. In
addition, Radu Rusu, creator of the
point cloud library (PCL), left Willow

IEEE ROBOTICS & AUTOMATION MAGAZINE

Garage to found the Open Perception
Foundation.
Impact and Legacy
Willow Garage created a number of important robot technologies. On the software side, ROS has become the
standard for controlling robots in academia, and is starting to make inroads
into industry. An effort called ROS Industrial is aimed at bringing ROS technology to industrial robot arms,
Rethink Robotics chose ROS for its new
dual-armed robot, Toyota has demonstrated a home robot running ROS, and
there are any number of robotics companies now shipping ROS-compatible
robots, or robots with ROS inside.
OpenCV and PCL both make significant tools available for robot perception,
in a ROS-compatible way. Through the
leadership of Ken Conley, contributions
to ROS were systematically documented through a series of blog posts on the
Web site WillowGarage.com in most

MARCH 2014

cases with corresponding videos on
YouTube. Most of the videos end with
the phrase "the code for this work is
open and available on ros.org."
On the hardware side, the PR2 continues to be used by graduate students to
make advances in robotics, especially at
a full-system level. There are approximately 50 PR2s in the world, in many of
the top robotics labs. The Texai remote
presence robot is now a commercial
product through Suitable Technologies'
Beam. The Platformbot prototype is
being commercialized through the Unbounded Robotics UBR-1 robot. Turtlebot, a low-cost platform for running
ROS, made available as open source
hardware, is on its second version and is
also well-represented in academia.
Although Willow Garage was clearly
more than a research lab, its research scientists made significant and sustained scientific impact. The company was well
represented in the major robotics conferences, both in published papers and in
organized workshops. Willow Garage scientists were charged with having scientific impact and community impact, as well
as impact on the Willow Garage products
(PR2, ROS, OpenCV, PCL, and so forth).
But as important as the robots, software, and research are, by far the most
important legacy of Willow Garage is
the people who came together there and
the relationships that will last for decades. At its peak, Willow Garage had
over 60 full-time employees, and approximately 100 people worked there
during its six and half years. There were
also many more visiting scholars and
interns who spent months at Willow
Garage. If you add up the number of
employees at the spin-off companies
mentioned above plus the team that
joined Savioke (my new company)
shortly after Willow Garage ended,
there are more people employed than
there had been at Willow Garage at its
peak. Eight Willow Garage employees
became first-time CEOs of robotics
companies in their next job after leaving
Willow Garage. We had set a goal of
creating a new industry, and this legacy
of new companies is a major step forward toward that goal.

http://www.WillowGarage.com

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2014