IEEE Robotics & Automation Magazine - December 2015 - 14

tc spotlight

Mobile Manipulation:
Toward Smart Manufacturing
By Máximo A. Roa, Dmitry Berenson, and Wes Huang

T

he goal of autonomous mobile
manipulation is the execution
of complex manipulation
tasks in both structured and
unstructured environments, possibly
in cooperation or close interaction
with human beings. This entails challenges in different fields: perception,
navigation, task and path planning,
control, error recovery, and human-
robot interaction. Each field is an area
of research on its own, but the special
challenge in mobile manipulation is to
obtain an integrated system that can
combine a large variety of hardware
and software components to increase
the tasks that the robot can perform,
while decreasing the dependency on a
priori information and increasing the
awareness of the robot to its current
situation. Challenges in the field arise
from the high-dimensional spaces associated to systems with high number
of actuators and sensors, uncertainty
in perception and execution of the
tasks, generality of the solutions required to manipulate objects in the
real world, and engineering and scientific complexity of the systems required for these tasks.
To foster the exchange of ideas in this
field, the IEEE Robotics and Automation Society (RAS) Technical Committee (TC) on Mobile Manipulation was
created in 2010. The activities of the TC
encompass research in the mentioned
areas, mostly with a focus on autono-

Digital Object Identifier 10.1109/MRA.2015.2486583
Date of publication: 10 December 2015

14

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

mous systems that execute complex
manipulation tasks. The scope of the TC
is not restricted to mobile manipulators
in industry but can also include applications in a large variety of scenarios, such
as homes, hospitals, or stores, including
tasks such as manipulation of industrial
parts in workcells, retrieval of objects in
warehouses, or assistance to patients in
healthcare facilities. In the short term,
these concepts could also apply to aerial
manipulation, for load transportation or
assembly of modular structures, or to
manipulation in space, a field where
manipulators have a large application
potential for maintenance, refueling, and
repairing of satellites. In fact, aerial and
spatial manipulation are frontier applications where the mobile base has six
degrees of freedom, in contrast to the
traditional mobile manipulators that
mount one or two arms on a base with
three degrees of freedom.
The last few years have seen several
mobile manipulators coming to market
(Figure 1), for example:
● Tiago by Pal Robotics: a single-arm
mobile manipulator with anthropomorphic hands and a lifting torso to
increase its vertical range [1].
● Mobile Baxter, by Rethink Robotics and
Clearpath Robotics: the dual-arm Baxter platform can be mounted on
Ridgeback, a specially designed mobile
base, to endow it with the mobility
required to explore large areas [2], [3].
● Fetch and Freight by Fetch Robotics: a
two-robot solution to provide efficiency in both ends of the spectrum,
with a single-arm mobile manipulator on a lifting torso, which feeds a

DECEMBER 2015

smaller and faster mobile base that
functions as a delivery cart [4].
Additionally, several industrial
robotics companies have also released
dual-arm manipulators to the market,
for instance YuMi from ABB or Nextage
from Kawada Robotics.
Traditional fixed-base robotic
manipulators have been ubiquitous in
large-scale factories for over 50 years, for
example, in the automotive industry.
Right now, the next niches to explore
with mobile manipulators are manufacturing at small and medium-sized companies, and in logistics. Typical examples
include the production of electronic
goods, or fulfillment in warehouse environments. Most of the robots required
for these applications should act and
move around people in an autonomous
way, so they would not require fences,
and therefore, need to consider human-
robot interaction to be successfully
deployed.
The archetypical example of logistics
is Amazon. The company owns gigantic
warehouses and has already automated
one portion of the workflow by using
mobile robots from Kiva Systems (now
Amazon Robotics) to bring shelves to
people in the picking and packing areas.
However, people still manually stock
the shelves, pick items from the shelves,
and pack orders into boxes for shipping.
Robots have considerable potential to
improve logistics operations but must
overcome the challenge of capably
manipulating a huge variety of products-large and small, heavy and light,
rigid and nonrigid-in densely packed
shelves and boxes.
Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2015
