Collaborative Pick and Place Using Indus

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Visión Electrónica Vol. 14 No. 1 (2020) • January • p.p.

72-78 • ISSN 1909-9746 • ISSN-E 2248-4728 • Bogotá (Colombia)

Visión Electrónica
Más que un estado sólido
https://doi.org/10.14483/issn.2248-4728

VISION ELECTRONICA

A CASE-STUDY VISION

Collaborative Pick and Place using industrial robots, simulation


and deployment
Simulación e implementación de proceso de Pick and Place colaborativo
con dos robots industriales
1 2
Camilo Alberto Pinzón-Quintero , Ángel Andrés Joya-Quintero , Felipe González-Roldán3 ,
4
Pedro Fabián Cárdenas-Herrera

información del artı́culo abstract


Historia del artı́culo: In recent times, a fraction of industrial robotics has been oriented to a collaborative
Enviado: 12/12/2019 approach between different devices to achieve a specific task. This paper describes
Recibido: 30/01/2020 the simulation and deployment of a metallic beam pick a place process using tow
Aceptado: 06/04/2020 industrial robots. At first, the tool design is presented; it used an electromagnet
attached to the robot end effector to lift the load correctly. The communication
Keywords: sequence between the robot’s controllers using the digital I/O modules is
End-effector design discussed. The software Robot Studio allowed the process simulation, in which
Industrial communications the performance was validated. All the tasks were tested in two ABB IRB 140
Industrial manipulator industrial robots.
Robot programming
Synchronized trajectories

resumen

En los últimos años la robótica industrial se ha orientado hacia la coordinación entre


diferentes manipuladores para realizar una determinada labor. En este sentido, el
presente documento expone la simulación e implementación de un proceso Pick and
Place de una viga metálica haciendo uso de dos robots industriales. Para tal fin,
Palabras clave: se diseñó una herramienta de electroimán para levantar apropiadamente la carga.
Diseño de efector final Además, se elaboró una secuencia de comunicación entre los dos robots haciendo
Comunicaciones industriales uso de los módulos de I/O digitales de los controladores, para ası́ sincronizar el
Manipulador industrial movimiento de los robots de manera sencilla. Finalmente, se implementó el proceso y
Programación de robots se verificó el funcionamiento correcto de todos los sistemas descritos anteriormente.
Trayectorias sincronizadas

1 BSc. in Mechatronic Engineering, Universidad Nacional de Colombia, Colombia. E-mail: [email protected]


2 BSc. in Mechatronic Engineering, Universidad Nacional de Colombia, Colombia. E-mail: [email protected]
3 BSc. in Mechatronic Engineering, MSc. in Industrial Automation, Universidad Nacional de Colombia, Colombia. E-mail:
[email protected]
4 BSc. in Electronic Engineering, Specialist in Industrial Automation, Universidad Pedagógica y Tecnológica de Colombia, Colombia. MSc.
in Industrial Automation, Universidad Nacional de Colombia, Colombia. MSc. in Automation and Robotics, Universidad Politécnica de Madrid,
Spain. Current position: professor at Universidad Nacional de Colombia, Colombia. E-mail: [email protected]
Cite this article as: C. A. Pinzón-Quintero, Á. A. Joya-Quintero, F. González-Roldán and P. F. Cárdenas-Herrera, “Collaborative Pick
and Place using industrial robots, simulation and deployment”, Visión electrónica, vol. 14, no. 1, pp. 72-78, january 2020.
https: // doi. org/ 10. 14483/ 22484728. 16364
73 C. A. Pinzón-Quintero, Á. A. Joya-Quintero, F. González-Roldán and P. F. Cárdenas-Herrera

1. Introduction tasks in a factory. In this way, a couple of robots that


performed separated roles in a production line can be
Industrial robots have played a vital role in the employed in a cooperative task, such as the handling
development of automated factories over the past 50 of a massive piece without purchasing new equipment.
years. The growth of robotic manipulators related to The processing of a rigid and heavy material, such a
industrial applications has been exponential, since 1961, metallic beam implies an additional overload on the
when Unimation was installed at the General Motors robot’s actuators [8]; to prevent a premature failure on
assembly plant in Detroit. According to the IRF [1], the robot, the load distribution among multiple robots
in 2017, more than 2 million robots had been installed is considered a feasible solution [9].
over the world. The number of industrial robots has
allowed diversification of its uses; at the beginning, the This paper aims to run a collaborative P&P process
main applications were welding, assembly, painting, and of a metallic beam using two industrial robots ABB IRB
machining. Even though nowadays, it is common to find 140 located in the LabSIR (Laboratorio de Sistemas
robots in such complex scenarios, like medical procedures Inteligentes Robotizados) at the Universidad Nacional
and collaborative work cells with human workers [2]. de Colombia. The process was implemented using the
controllers I/O modules for each robot. Also, a specific
The process of move an object from a first work work sequence was created to allow the synchronization
station to a second work station, being both stations of the robot’s trajectories, without damages in the
located in the robot’s workspace, is commonly known moving piece nor the robot’s mechanisms.
as pick and place (P&P) [3]. It is worthy of mentioning
that a P&P process implies a pose change, which means a This document is organized into three sections. At
position and orientation modification. In some industrial first, an introduction of automated and robotic solutions
applications to solve the handling of heavy pieces, two is stated, it is emphasized the need for collaborative
approaches are employed. The first one consists of tasks in the modern industry, especially in the Colombian
the use of high load manipulators, and secondly, the context. In the second part, the methods used to
incorporation of multiple robots that executes tasks in a accomplish the collaborative P&P are described, starting
coordinated way. with the end effector design, followed by the simulation
process done in Robot Studio and finishing with the
The utilization of heavy-duty robots is the most deployment in the actual robots. The third section
common way to solve the handling of heavy pieces. discusses the results obtained, and then the conclusions
Nevertheless, in recent times multiple industrial robot and future work are mentioned.
manufacturers have developed tools that allow the use
of one or more robots in a collaborative way. Some
2. Methods
recognized applications are Multimove by ABB [4]
or RoboTeam by Kuka [5]. These tools enable the
A set of steps was proposed to accomplish the
synchronized movement of multiple robots so that they
simultaneous collaborative P&P using two industrial
can share a common reference frame; therefore, they can
robots. Firstly, the tool design was performed; its
work on the same piece. It is relevant mentioning the
calibration was made through two different methods to
main difference among coordinated and synchronized
ensure its functionality as an end effector. Subsequently,
movement; the first one has to do with the reference
the robot’s motion and communication routines were
system that a set of robots use for the task execution
programmed in Robot Studio, which allowed a complete
process [6]; while in the case of the second one, the
simulation of the process. As the final step, the
movement of the robot is done at the same time [6].
simulation results were deployed and tested on the real
industrial manipulators.
Specialized applications such as Multimove and
Roboteam requires specific hardware and software
modules, which implies additional expenses in 2.1. Tool design
comparison to the cost of a traditional robotic solution.
Colombian industry still is at an early stage regarding 2.1.1. Mechanical design and components selection
automated and robotic stations [7]. Therefore, companies
interested in robotics need technological solutions at the The main design parameters for the tool were the tool
lowest cost possible. Also, robotic systems should be material and the piece weight. The selection process was
as flexible as possible; so, they can be used in different accomplished by using a specialized catalog; finally, an

Universidad Distrital Francisco José de Caldas - Facultad tecnológica


Collaborative Pick and Place using industrial robots, simulation and deployment 74

electromagnet that met all the requirements was chosen, Two approaches may be used: thought software or using
the component has the following technical specifications: a physical procedure. The first one requires the software
Robot Studio, in which the CAD model of the tool
• Brand: Security Home.
is imported for its virtual calibration (Figure 2). Two
• Power supply: 12-14 VDC coordinate frames are created, one located on the tool
base and another on the TCP (Tool Central Point). Then
• Load capacity: 180 lbf (max). both structures are linked through a transformation
• Dimensions: 72x31x27 mm. matrix, which is calculated by the data provided. The
physical approach requires that the tool was mounted
An L shape plate was designed to couple the at the robot flange; the Teach Pendant offers a series
electromagnet to the robot’s endpoint. The longest part of options to calibrate the robot tool. According to the
of the plate is used to connect the robot’s flange to the electromagnet geometry, the most suitable one was the
tool trough M6 screws [10]; in the shortest side of the mode with 3 points and one axis; therefore, that method
plate, the electromagnet is located, which is attached was selected. The electromagnet center is located at the
via a couple of M5 screws. In Figure 1 is presented the same position with three different orientations. Next,
machine tool design. using a vertical axis that crossed the fixed point, the tool
approach axis is aligned with it, and an approximation
process is made to the fixed point. The robot controller
Figure 1: Machine tool design intended for metallic
automatically computes a transformation matrix that
pieces gripping.
represents the tool.

Figure 2: Virtualized tool calibrated in Robot Studio.

Source: own
The P&P requires two processes related to the
Source: own
piece, attachment, and release; the first one was solved
by using the electromagnet and the proposed tool
as well, the second process demanded an additional Table 1 shows information related to the tool
component called antiremanent module, whose function calibration process. The position data in both procedures
is to demagnetize the tool, this effect was reached in an is relatively comparable. On the other hand, orientation
average time of 0.5 s. results show a considerable difference, since the virtual
process is no susceptible to human error, the data
2.1.2. Tool calibration provided from this method was chosen. The most likely
reason for the orientation difference has to do with the
With the tool assembled and mounted in the axis aligning process in the physical approach, which is
manipulator, the calibration process was carried out. human vision guided, hence highly subject of error.

Visión Electrónica Vol. 14 No. 1 (2020) • January • p.p. 72-78 • ISSN 1909-9746 • ISSN-E 2248-4728 • Bogotá (Colombia)
75 C. A. Pinzón-Quintero, Á. A. Joya-Quintero, F. González-Roldán and P. F. Cárdenas-Herrera

Table 1: Comparison of the tool data calibration results. Position (X, Y, Z) expressed in (mm); orientation expressed
in quaternions (Q1, Q2, Q3, Q4).

Source: own

Table 2: Motion parameters required to program an ABB industrial robot.

Source: adapted from [4].

Figure 3: Programmed trajectories for each robot.

Source: own

2.2. Robot programming and simulation in Robot A set of specialized coordinate systems called
Studio workobjects were located at the same point in the beam
for each robot. The primary purpose of the workobjects is
The robot’s programming and simulation started by to guide the targets that control the robot’s trajectories.
locating two ABB IRB 140 manipulators in the software It is worth mentioning that workobjects are easily
Robot Studio, according to the layout in the LabSIR. changeable. If they are modified, all data associated
An additional extruded prism was located to emulate with them is modified as well, so much work can be
the beam with which the process was going to be done. saved when path changes are required.

Universidad Distrital Francisco José de Caldas - Facultad tecnológica


Collaborative Pick and Place using industrial robots, simulation and deployment 76

With the proper references specified, the robot’s


trajectories were created. In Robot Studio, the waypoints Figure 4: Communication sequence between industrial
that each robot had to pass through were defined. Then robots used in this application.
those points were linked by a specific type of motion.
Table 2 shows the required motion parameters for the
ABB industrial robots.

The motion sequences programmed for each robot are


described as following:

• Approaching to the beam: MoveJ, speed v200, zone


fine (the TCP passes precisely through the target).
• Beam lifting: MoveL, speed v600, zone z10 (the
TCP passes to a max. distance of 100mm to the
target).
• Beam translation: MoveL, speed v1000, zone z10.
• Beam orientation change: MoveL, speed v600, zone
10. This motion rotates the beam 90◦ without
translation.
• Beam dropping: MoveL, speed v600, zone fine.
• Return to home: MoveJ, speed v200, zone fine. Source: own

Figure 3 shows the created trajectories for each 2.3. Deployment in industrial robots
industrial robot. All processes elaborated on the
station mode of Robot Studio (visual interface) were After the simulation validated the results, the motion
synchronized to a RAPID module, which contained routines were implemented in the industrial robots
the programming commands to be executed in the ABB IRB 140. In section 2.1 was mentioned that the
robot controller. It is worth highlighting the fact that robot’s tool used an antiremanent module to control its
the RAPID module stores all information related to magnetization; this component has an integrated buzzer
the motion. It includes the tool calibration data, the to indicate when it is enabled. The antiremanent module
reference frames, workobjects, targets, and trajectories. had a push-button as an original control method. A relay
Also, this information is used to create the motion replaced this button, which can be activated from the
routine and the communication protocol between the two robot controller using digital output; in this way, the
robots. The communication sequence was implemented magnetization process was controlled straightly by the
using the available digital Input/output module in the robot.
robot’s controller panel.
With the connections diagram done (Figure 5), the
In general, the communication sequence consisted of I/O pins on each IRC5 controller were assigned. Tables 3
each robot enables a digital flag when the initial pose and 4 present the pins assignation and the corresponding
was reached. At the same time, each robot waited until functions.
the other robot arrived at the initial pose, and then start
Table 3: Input/output assignation for robot controller
the motion. Figure 4 shows the graphical representation
number 1.
of the communication sequence. Although the two robots
arrive at different times to the initial pose, each one waits
to receive the other robot signal to start the process.

With each robot motion routine programmed,


a simulation in Robot Studio allowed a validation
procedure. Targets and motion parameters were adjusted
in the same way for both robots; the simulation showed
that the entire process was continuous, and it lasted
approx. 12 s. Source: own.

Visión Electrónica Vol. 14 No. 1 (2020) • January • p.p. 72-78 • ISSN 1909-9746 • ISSN-E 2248-4728 • Bogotá (Colombia)
77 C. A. Pinzón-Quintero, Á. A. Joya-Quintero, F. González-Roldán and P. F. Cárdenas-Herrera

Figure 5: Tool connections diagram. Table 4: Input/output assignation for robot controller
number 2.

Source: own Source: own.

Figure 6: Real implementation using two ABB industrial robots.

Source: own

3. Results process execution. As a result, no delay in the motion


was observed, nor the piece suffered any damage.
The total time for the test using the real robots was
12.15 s. The approach motion lasted 4.24 s, 1.14 s for 4. Conclusions and discussion
the communication sequence, 2.19 s of robot motion,
and 3.15 s for the trajectory to the initial pose. The By utilizing the digital input/output modules,
simulation time in Robot Studio was 12 s; therefore, the included in most of the industrial robot’s controllers,
difference is barely existent. (See Figure 6). the implementation of a collaborative P&P task was
possible. Processes that require synchronous motion can
It is worth mentioning that by using the digital be done with ease, using basic communication sequences
input/output module in the robot’s controller, the between multiple manipulators.
motion synchronization between the two robots was
satisfactory. Although the robots arrived at different The previous idea served as the base to a real task,
times, the defined communication sequence allowed the in which two robots handled a metallic beam. Usually,
simultaneous movement. A tiny time gap was present this task would have been required either a bigger robot
due to time that the robots use to send and receive the or the use of specialized software.
digital signals; nevertheless, this time did not affect the

Universidad Distrital Francisco José de Caldas - Facultad tecnológica


Collaborative Pick and Place using industrial robots, simulation and deployment 78

A small-time difference was observed by comparing [4] ABB, “IRC5 Industrial Robot Controller”. [Online].
the times obtained in both the simulation and the real Available at : https://new.abb.com/products/
test. This time gap is the result of the procedures robotics/controllers/irc5
executed inside the robot’s controller; nevertheless, it
did not affect the actual task in any sense. [5] KUKA, “KUKA.RoboTeam”. [Online].
Available at: https://www.kuka.com/es-es/
According to the obtained results, a comparison productos-servicios/sistemas-de-robot/
between the implemented solution and a task using the software/tecnolog%C3%ADas-transversales/
ABB Multimove option would be useful to identify the kuka_roboteam
limitations of the approach developed in an industrial [6] A. Blomdell, I. Dressler, K. Nilsson, and A.
context. Also, the use of different communication Robertsson, “Flexible Application Development and
protocols between robots is proposed to determine if High-performance Motion Control Based on External
better motion synchronization can be achieved. Sensing and Reconfiguration of ABB Industrial
Robot Controllers”, in IEEE Int. Conf. Robot.
Autom., pp. 62–66, 2010.
Acknowledgments
[7] J. E. Pinzón-Varela, “Diseño de un sistema
Special thanks to the Universidad Nacional robotizado que permita paletizar el embalaje de
department of mechanical and mechatronic engineering, jugos en presentación de cajas tetra pack de 200ML,
to the Laboratorio de Sistemas Inteligentes Robotizados en la empresa de Productos Alimenticios Alpina
(LabSIR) and all its personnel for having helped in the Cauca”, thesis, Universidad Autónoma de Occidente,
development of the present work. An acknowledgment to Colombia, 2017.
Brian Avila for all his contributions aided to finish this
[8] I. B. Chelpanov, and S. N. Kolpashnikov, “Problems
paper successfully.
with the mechanics of industrial robot grippers”,
Mech. Mach. Theory, vol. 18, no. 4, pp. 295–299,
References 1983. https://doi.org/10.1016/0094-114X(83)
90122-2
[1] IFR International Federation of Robotics,
“Executive Summary World Robotics 2017 Industrial [9] K. Tai, A.-R. El-Sayed, M. Shahriari, M.
Robots”, pp. 23–24, 2017. [Online]. Available at: Biglarbegian, and S. Mahmud, “State of the
https://ifr.org/downloads/press/Executive_ Art Robotic Grippers and Applications”,
Summary_WR_2017_Industrial_Robots.pdf Robotics, vol. 5, no. 2, p. 11, jun. 2016.
https://doi.org/10.3390/robotics5020011
[2] B. Siciliano, and O. Khatib, “Springer Handbook
of Robotics”, Springer, 2007. https://doi.org/10. [10] ABB, “Product specification Articulated
1007/978-3-319-32552-1 robot IRB 140”, p. 19, 2004. [Online].
Available at: https://library.e.abb.com/
[3] C. V. Doan, “Pick and Place Machine”, U.S. Patent public/603383cd889d40598d1d2bf993e0f503/
5564888A, September 27 1993. 3HAC9041-1_rev6_en_library.pdf

Visión Electrónica Vol. 14 No. 1 (2020) • January • p.p. 72-78 • ISSN 1909-9746 • ISSN-E 2248-4728 • Bogotá (Colombia)

You might also like