14-02-2013, 01:53 PM
Accelerometer-Based Control of an Industrial Robotic Arm
Accelerometer-Based Control.pdf (Size: 456.75 KB / Downloads: 57)
Abstract—
Most of industrial robots are still programmed
using the typical teaching process, through the use of the robot
teach pendant. In this paper is proposed an accelerometer-based
system to control an industrial robot using two low-cost and
small 3-axis wireless accelerometers. These accelerometers are
attached to the human arms, capturing its behavior (gestures
and postures). An Artificial Neural Network (ANN) trained
with a back-propagation algorithm was used to recognize arm
gestures and postures, which then will be used as input in the
control of the robot. The aim is that the robot starts the
movement almost at the same time as the user starts to perform
a gesture or posture (low response time). The results show that
the system allows the control of an industrial robot in an
intuitive way. However, the achieved recognition rate of
gestures and postures (92%) should be improved in future,
keeping the compromise with the system response time (160
milliseconds). Finally, the results of some tests performed with
an industrial robot are presented and discussed.
INTRODUCTION
ROGRAMMING and control an industrial robot
through the use of the robot teach pendant is still a
tedious and time-consuming task that requires technical
expertise. Therefore, new and more intuitive ways for robot
programming and control are required. The goal is to
develop methodologies that help users to control and
program a robot, with a high-level of abstraction from the
robot specific language. Making a robotic demonstration in
terms of high-level behaviors (using gestures, speech,
manual/human guidance, from visual observation of human
performance, etc.), the user can demonstrate to the robot
what it should do [1]-[5].
In the robotics field, several research efforts have been
directed towards recognizing human gestures, recurring to
vision-based systems [6], [7], motion capture sensors [2],
[4], or using finger gesture recognition systems based on
Manuscript received March 15, 2009. This work was supported in part
by the European Commission’s Sixth Framework Program under grant no.
011838 as part of the Integrated Project SMErobotTM, and the Portuguese
Foundation for Science and Technology (FCT), grant no.
SFRH/BD/39218/2007.
Pedro Neto is a PhD student in the Industrial Robotics Laboratory -
Mechanical Engineering Department of the University of Coimbra, POLOII,
Pinhal de Marrocos, 3030-788, Coimbra, Portugal; (e-mail:
pedro.neto[at]robotics.dem.uc.pt).
J. Norberto Pires is with the Industrial Robotics Laboratory -
Mechanical Engineering Department of the University of Coimbra, POLOII,
Pinhal de Marrocos, 3030-788, Coimbra, Portugal; (e-mail:
jnp[at]robotics.dem.uc.pt).
A. Paulo Moreira is with the Institute for Systems and Robotics (ISR) -
Department of Electrical and Computer Engineering of the University of
Porto, Rua Dr. Roberto Frias, 4200-465, Porto, Portugal; (e-mail:
active tracking mechanisms [8]. Accelerometer-based
gesture recognition has become increasingly popular over the
last decade. The low-moderate cost and relative small size of
the accelerometers make it an effective tool to detect and
recognize human body gestures. Several studies have been
conducted on the recognition of gestures from acceleration
data using Artificial Neural Networks (ANNs) [9], [10],
[11]. However, the specific characteristics of an industrial
environment (colors, non-controlled sources of light,
infrared radiation, etc.), the safety and reliability
requirements, and the high price of some equipment has
hampered the deployment of such systems in industry.
Given the above, the teach pendant continues to be the
common robot input device that gives access to all
functionalities provided by the robot (jog the manipulator,
produce and edit programs, etc.). In the last few years the
robot manufacturers have made great efforts to make userfriendly
teach pendants, implementing intuitive user
interfaces such as icon-based programming [12], color touch
screens, a 3D joystick (ABB Robotics), a 6D mouse (KUKA
Robot Group) [13], or developing a wireless teach pendant
(COMAU Robotics). Nevertheless, it remains difficult and
tedious to operate with a robot teach pendant, especially for
non-expert users.
In this paper is proposed an accelerometer-based gesture
recognition system to control an industrial robot in a natural
way. Two 3-axis wireless accelerometers are attached to the
human arms, capturing its behavior (gestures and postures).
An ANN system trained with a back-propagation algorithm
was used to recognize gestures and postures. Finally, several
tests are done to evaluate the proposed system. The results of
the performed tests are presented and discussed.
SYSTEM OVERVIEW
System Description
The demonstration cell (Fig. 1) is composed of an
industrial robot MOTOMAN HP6 equipped with the NX100
controller, two 3-axis wireless accelerometers to capture
human hand behaviors, and a computer running the
application that manages the cell.
The 3-axis accelerometers (ADXL330, Analog Devices)
are physically rated to measure accelerations over a range of
at least +/- 3g, with a sensitivity of 300 mV/g and sensitivity
accuracy of 10%. The accelerometers communicate with the
computer via Bluetooth wireless link, reporting back data at
100 Hz (Fig. 2).