View metadata, citation and similar papers at [Link].
uk brought to you by CORE
provided by Journal of Student Science and Technology (The Foundation for Student Science and...
ARTICLES
DESIGN, PROTOTYPING AND PROGRAMMING OF
A BRICKLAYING ROBOT
Aguiar, M. L., Behdinan, K.
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario M5S 3G8, Canada
ABSTRACT
This paper presents the development of a Bricklaying Robot capable of building entire walls. This project was
motivated by opportunity of automation in construction, which remains costly and inefficient. Several studies have
described specialized robot for masonry works, most of them incorporating the human arm concept that requires
complex programming and low productivity. In order to improve this model, an innovative concept of a short arm
assembled into a lift platform was introduced in this paper. The robot was modeled in Solidworks®, followed by
motion study and dynamic analysis to optimize the model. A prototype in 1/4 scale was built to demonstrate
its feasibility, detecting and correcting flaws. The prototype was tested by programming servo motors using
Arduino UNO hardware and C ++ code. Finally, robot kinematics was analyzed in a construction site scenario.
Concordance between the virtual simulation and the experimental prototype results demonstrated the functionality
and effectiveness of the proposed design. The invention of this Bricklaying Robot will represent a technological
advancement in developing new mechanisms and codes, which may be responsible for increasing productivity and
reducing risks of masonry construction. Future analysis of global bricklaying market might be conducted to prove
its commercial viability.
Cet article présente le développement d’un robot de maçonnerie capable de construire des murs entiers. Le
projet est motivé par une grande possibilité pour l’automatisation dans le bâtiment, un domaine qui reste malsain,
coûteux et inefficace. Plusieurs recherches discutent des robots spécialisés pour les travaux de maçonnerie ;
la plupart de ces robots utilisent le modèle du bras humain, une conception qui nécessite la programmation
complexe et présente une perte de productivité. Cet article a pour but d’améliorer le modèle courant en explorant
une nouvelle conception d’un bras court, assemblé comme une plate-forme élévatrice. Ce robot a été conçu en
Solidsorks®, et une étude de mouvement et une analyse dynamique ont été menées pour optimiser le modèle.
Un prototype à l’échelle 1:4 a été conçu pour démontrer la faisabilité du robot, tout en détectant et corrigeant ses
défauts. Le prototype a été évalué par la programmation des servomoteurs, en utilisant du matériel informatique
d’Arduino UNO et le langage de programmation C ++. Finalement, la cinématique du robot a été analysée
en un scénario de chantier de construction. La fonctionnalité et l’efficacité de la conception proposée ont été
démontrées par la concordance entre la simulation virtuelle et les résultats du prototype expérimental. L’invention
de ce robot de maçonnerie constituera une avance technologique dans l’étude et le développement de nouveaux
mécanismes et des codes informatiques, qui permettront d’améliorer la productivité et réduire les risques de
maçonnerie. Une analyse de marché mondial de maçonnerie peut être menée à l’avenir pour vérifier la viabilité
commerciale du robot.
INTRODUCTION
Several industries have undergone a complete Scientific literature suggests that masonry work is
automation process in the last century, especially in the often related to human health and diseases. In 1999,
agriculture, automotive, food, logistics, and machinery a study assessing the occurrence of musculoskeletal
industries. However, unlike the previously mentioned disorders observed that 38% of the bricklayers in
industries, the construction industry is going through a the Netherlands reported back complaints[2]. Many
very slow process of automation. While the production external factors to the bricklayer him/herself can
of industrial robots is increasing up to millions of units, have a considerable effect on productivity; Ultimately
there are only hundreds of robots in masonry works however, the average bricklaying rate of one human is
that are produced [1] . 140 bricks/hour [3] . In addition to all the disadvantages
DOI: 10.13034 / JSST-2015-032 THE JOURNAL OF STUDENT SCIENCE AND TECHNOLOGY 2015 VOL 8 ISSUE 3 27
of manual bricklaying, 1993 studies reported a lack MATERIALS AND METHODS
of skilled labor in the German masonry construction Definition of the Robot’s Concept and Tasks
industry[4]. Several studies observed a high rate
of inefficiency in the masonry industry of different The model was brainstormed based on a lift platform
countries. For example, studies reported inefficiencies inside four-tubes column. An automatic gripper, which
at approximately 68%, 50%, 40%, 7%, and 2% for grabs each brick delivered by the flow rack and lays
firms in Spain[5], Canada[6], Portugal[7], Greece[8], and it on the wall after injection of mortar by nozzles, are
China[9] respectively. assembled on the lift platform. A column must be
assembled onto a mobile base with wells and tracks.
It was proven that automation increases productivity
and reduces the risk of work disorders in highly Design of Structure and Mechanism in CAD Software
repetitive and physically strenuous tasks. Robotics The 3D model parts were designed through usual
have been consolidated with the most important features of extrusion, revolve extrusion, and cut of
technological advancement in the automation process. Solidworks®. Tubes and steel sheet dimensions were
In Israel, a mobile robot for building interior walls based on standard commercial brochures. Steel ASTM
was developed [10]. Meanwhile, a group of European A36 and AISI 1020 were defined as material for most
engineers developed a robot called ROCCO, which of the components from Solidworks Materials Library.
used an algorithm to plan, move its arm, and work Then, an assembly document was created joining all
on complex bricklaying situations[11]. For the past parts through regular and mechanical mates. Finally,
2 decades, Increasingly advanced electronics a technical drawing was generated from the assembly
sensors are being used in building equipment[12]. document with all dimensions and list of material
Quality control and Tool Center Point (TCP) calibration necessaries to build the robot.
are achieved by using vision systems, laser beacons Kinematic Analysis
and ultrasonic sensors[13]. Automated laying of
Structure was tested in Solidworks® to verify its
conventional clay bricks and application of bonding
behaviour during bricklaying process. Movements
material is investigated in the United Kingdom by
analysis were conducted using the tool Motion Study
means of a stationary gantry type robot[10].
to verify if the short arm was the correct length and
Generally, the bricklaying mechanism involves movement range while it grabbed each brick and
the concept of the human arm, as discussed in moved it on the wall. Also, to set up the velocity of base
literature [4, 10, 11, 13]. This concept is based on a long movement, it was necessary to adjust the angular
arm with a vacuum gripper at the upper end, which velocity of the arm. A 10-second video was recorded
is assembled onto a mobile base. Thus, this paper from the software Motion Study for analysis.
presents an innovative mechanism that moves each
Prototyping of 1/4 scale robot
brick faster and injects mortar at the same time. Also,
this project aimed to design, prototype, and program A prototype in 1/4 scale of the original concept
a new concept of short arm that is assembled onto a dimensions was manufactured using aluminum
lift platform. It was necessary to calculate usual loads beams, steel sheets, plastic components, three DC
in the long arm concept to find out why they have motors of 12V, and one servo motor of 5V.
heavier structures, resulting in over consumption of Programming of movements of prototype
energy, high complexity of positioning programming
The movements of each motor were programmed using
and low productiveness. The new mechanism was
Arduino UNO® board and C++ code. Programming is
tested, while the results were compared to the long
based on state machine concept to control base and
arm concept to confirm its feasibility and whether
gripper movements at the same time.
the design allows for manufacturing of a lighter and
more portable structure, which would have higher Experimental tests and mechanism analysis
productivity in masonry works. Tests were conducted with 1/4 scale bricks to measure
velocity and acceleration of prototype. The movements
28 2015 VOL 8 ISSUE 3 LA REVUE POUR LES ÉTUDIANTS EN TECHNOLOGIE ET SCIENCES DOI: 10.13034 / JSST-2015-032
of the robot were recorded in videos and analyzed in
2X slow motion using basic video software such as
Movie Maker®. The results were compared to those
from robots in observed literature.
RESULTS
The Design of a Final Structure and Mechanism
The 3D modeling of the Bricklaying Robot resulted in
a model consisting of a four-tubed column that can be
extended from 1.60 m to 3.00 m. Lateral dimensions
were reduced to 500 mm x 500 mm, allowing the
robot to pass through doors. The total weight of
the structure was around 70 kg to ensure that the
structure can be carried by a single worker (Figure 2).
Gripper Mechanism Was Conceptualized and Evaluated
The top end of the arm has an automatic gripper Figure 1:
which was designed and analyzed in terms of forces
Usual concept of robot for masonry works which
and resistance during the process of picking up and
design is based on human arm. Average weight of a brick
moving each brick. The final mechanism design did and estimated weight of manipulator components were
not have any electronic sensor or actuator (Figure 3). used in the calculus of momentum on the base articulation.
Equation 1 represents the minimum force applied by
each side of the gripper (GF) to hold one brick:
2.7 kg x 9.81 m
s2
w
GF 2 GF 2 GF 22.7 1
µ 0.6
Where, an average red clay brick (4” x 8” x 2 1/4”) weighs
2.7 Kg and the friction coefficient (µ) between a rubber
and brick surface (adopting concrete) is approximately
equal to 0.6. Therefore, the GF must be ≥ 22.07 N to
assure that LF=W/2. With this result (∑Fy=0) it was
observed that the gripper mechanism was capable of
holding one brick safely during movement.
The Prototype Was Built According to the 3D Model
The prototype in 1/4 scale from original dimensions
was built using a technical drawing from Solidworks®
(Figure 5). The materials that constituted the majority
of the prototype included aluminium tubes, bolts,
beams, steel sheets, and plastic components. It was
not necessary to weld any structure. All parts were
assembled using fasteners. Two gear motors and 1
servo motor were responsible for the movements. Figure 2:
One Arduino UNO board and 4 batteries of 9V each Proposed design for innovative Bricklaying Robot is
were responsible for controlling the motors. more compact. 3D model was generated in Solidworks
through part modeling and assembly. It was simulated
The Results of Kinematic Analysis using Motion Study tool and optimized until this final
Operation of the prototype combined different design, which is more portable, lighter, cheaper, quicker,
movements sequenced for moving and positioning and more efficient than long arm bricklayer concept.
DOI: 10.13034 / JSST-2015-032 THE JOURNAL OF STUDENT SCIENCE AND TECHNOLOGY 2015 VOL 8 ISSUE 3 29
Figure 3:
Automatic gripper mechanism is able to grab and hold a brick. (a) Gripper approaches over the brick; mechanical
sensor starts touching its upper face with sensor force (SF) and closes the holders. (b) Free body diagram where LF is
the frictional force due gripper force (GF) and roughness between contact surfaces (µ), W=2xLF. (c) Releasing movement
occurs after higher contact force on mechanic sensor during pressing the brick on the wall.
Figure 4:
Graph describing the programming of robot’s movements. The three motors are programmed together using only one
control hardware and one code based on State Machine principle. The movements were synchronized according the time
as the graph presents. While base DC motor moves constantly forward, gripper servo motor moves in an oscillatory way.
30 2015 VOL 8 ISSUE 3 LA REVUE POUR LES ÉTUDIANTS EN TECHNOLOGIE ET SCIENCES DOI: 10.13034 / JSST-2015-032
Figure 5:
1/4 scale prototype working on construction site
scenario. (a) Prototype was built according technical
drawing dimensions divided by 4. Structure is made from
aluminum. (b) Lift platform, gripper mechanism, Arduino
UNO board and mortar pump are presented.
Figure 6:
Frames sequence of one brick lying during 1.5 seconds.
The first image shows the arm approaching on the upper
face of the brick. The second frame presents the gripper
applying force and holding the brick. The brick moving is Figure 7:
shown by the third frame and the fourth frame is the final
movement of the gripper, when it releases the brick on its Flowchart describing the sequence of operation of the
final position on the wall. robot. Bricklaying Robot must be the most automatized as
possible. Steps in gray are represented by human actions
and in white by robot action. Hard and repetitive tasks are
clearly done by robot.
each brick. While the base moved straight forward, mortar pump; this graph also guided the process of
the bricks slid down and were picked up by the the robot’s programming.
gripper. Mortar was injected by nozzles onto bricks Test Outcomes from Experimental Prototype
in the wall and the gripper lay each brick over the
mortar layer. After each tier was completed, the base Figure 6 displays the frames of the recorded video
stopped moving forward, while mortar pump paused during tests with the Bricklaying Robot prototype.
during 1.5 seconds, and the platform was raised The average frequency of work measured in the first
to the equivalent height of 1 brick (2 1/4”). Then, test was approximately 1 brick per 1.51 seconds. The
the robot started moving backward, continuing the information below was measured from prototype tests
bricklaying function. The previewed frequency of after timing calibration.
the machine was 1 brick per 1.5 seconds. Figure 4 • Arm displacement: Backward = 59.10°; Forward =
describes the sequence movements of each motor 69.39°; Total = 128.49° = 2.24 rad
of time (seconds) versus electrical signal, except the Arm timing: Backward = 820 ms; Forward = 810 ms;
DOI: 10.13034 / JSST-2015-032 THE JOURNAL OF STUDENT SCIENCE AND TECHNOLOGY 2015 VOL 8 ISSUE 3 31
Total= 1630 ms = 1.63 s amount in degrees of freedom in a multi articulated
Arm speed = 78.82°/s = 1.3742 rad/s arm results in complexity of programming. In addition,
• Base speed = 41.3 mm/s = 0.0413 m/s they are very expensive and commercially unviable
Each cycle is being completed within 67.31 mm for small and medium-sized constructors.
(41.3 mm x 1.63 s = 67.31 mm). Considering The results from 3D modeling presented an innovative
that each prototype brick is 67.00 mm long, the and simple design which solves many of the difficulties
linear space for completing one cycle is sufficient of masonry construction that preclude the use of robots,
according these measurements. such as mortar injection on the wall during bricklaying,
• Elevator displacement: 20 mm size reduction, simplification of movements, and
Elevator delay: 1.52 s operating facilitation. Kinematics analysis proved that
Elevator speed: 13.2 mm/s = 0.0132 m/s expected movements are possible to be executed by
the mechanism. Speeds were defined and tested to
The prototype confirmed that the gripper mechanism achieve the maximum productivity without quality loss.
worked the same way it was designed; that applied A prototype was built and programmed in a relatively
force was enough to hold and move a brick. Dynamic easy way, proving that the concept is viable. The first
analysis on the computer resulted in well dimensioned design of mortar pumping system did not work as
components that can work under the required loads expected because its pressure loss due gaps between
and speeds. Programming of prototype movements screw and tube wall. The pump system does not work
resulted in a refined C++ code that was based on State for high viscosity fluids such as mortar or concrete,
Machine. Tests demonstrated that it is possible to have and because of that, it was redesigned. The second
a high level of automation of the bricklaying process concept of pump is based on a peristaltic pump with
with low human interference. However, data loading, a flexible hose made from latex that generates the
initial positioning of the robot, and refilling of materials pressure required for mortar pump. The experimental
are the only tasks required by humans (Figure 7). tests with the prototype observed the automatic
DISCUSSION gripper mechanism working at a rate of 1 brick per 1.5
seconds, or up to 2400 bricks per hour upon assuming
A bricklaying Robot must be light, mobile, easy to
a non-stop operation. When comparing prototype
operate, compact, precise, quick, and commercially
production rate to the average bricklaying rate of one
viable. Based on the robots presented by literature, it
human (140 bricks/hour), this projects suggests that
is possible to conclude that they have many physical
the robot is time efficient, low energy consuming and
features to improve upon in order to achieve the
cost for construction.
previously mentioned qualities. Although they are
mobile and accurate, they require a more robust The Bricklaying Robot is presented as a real
structure, which makes them heavy and large. The competitive equipment compared to the existing
concept based on a long arm produces a considerable robots available in the markets. Due to design
distance between the main load (brick) and base improvements, it was proved that the shorter arm
articulation, causing high moment load (Figure 1). It can lay bricks with the same precision of the long
is possible to state that the overload on articulations arm concept, with the advantage of being quicker
due high momentum causes high angular inertia and capable of injecting the mortar at the same time.
values, which obligates designers to reinforce the Concordance between virtual simulation and the
arm structure and limited the speed of operation. experimental prototype proved the feasibility of the
Due to the extra weight of arm structure, in most of innovative automatic gripper. The structure is easier
cases, they are equipped with hydraulic or pneumatic to build and the programming of its movements are
actuators that are characteristically slow (Figure 1). less complex compared to the usual concept. With all
Another disadvantage to these systems is that they these design improvements, it is possible to increase
require the pump and valve installation of pumps that considerably the productivity, the efficiency, and
take up space and add additional weight to the robot. reduce cost, accident risks and waste of masonry
In the case of electronic servo motors installation, the construction, turning it more sustainable.
32 2015 VOL 8 ISSUE 3 LA REVUE POUR LES ÉTUDIANTS EN TECHNOLOGIE ET SCIENCES DOI: 10.13034 / JSST-2015-032
FUTURE DIRECTIONS V Volts
The innovative mechanism was tested and worked vs. Versus
as designed, except the mortar pumping system. W Weight Force
The next directions for the Bricklaying Robot will
be defining the correct design for mortar pumping KEY WORDS
system based on peristaltic pump; this concept will Bricklaying Robot; Masonry Construction; Automation;
be developed and tested. The loads on articulations Mechanical Design; Prototyping
and structures need to be evaluated to determine
the correct shape, materials, bolts, bearings and
ACKNOWLEDGEMENTS
manufacturing processes. To proceed with this This work is supported by the Brazilian Government
evaluation, the Working Model 2D software will be with funds of Conselho Nacional de Desenvolvimento
used. The installation of sensors for closed-loop Cientifico e Tecnologico (CNPq). Programming of
system is very important for positioning scanning Arduino UNO board used in prototype was conducted by
of the prototype and quality improvement of the Isaías Junior, Electrical Engineering student. This project
bricklaying process. After assuring that prototype is was assisted by Professor Kamran Behdinan and master
working and injecting mortar as it is supposed to do, student Yasaman Saeedi at Department of Mechanical &
all components will be listed and budgeted to know Industrial Engineering of University of Toronto.
the final cost for manufacturing an original scale REFERENCES
prototype; this is essential because it determines the
commercialization feasibility. 1. Hasegawa, Y. Current Status and Key Issues for
Construction Automation and Robotics in Japan.
ABBREVIATIONS ISARC, 1996, Volume 13.
µ Friction Coefficient 2. Arbouw. Map of the Construction Industry ’98/’99
2D Two Dimensions [in Dutch]. Amsterdam, Netherlands, 2000.
3D Three Dimensions 3. Whitehead, B. Productivity in Bricklaying. Building
Science Journal, 1973, Volume 8, Pages 1-10.
AF Arm Force
4. Bitschow, G; Dalacker, M; Kurz, J; Zeiher
ASTM American Society for Testing and
J. A Mobile Robot for On-Site Construction of
Materials
Masonry. Stuttgart, Germany, 1996, Page 1701.
CAD Computer Aided Design
5. Kapelko, M; Oude Lansink, A. Technical
DC Direct Current Efficiency and its Determinants in the Spanish
GF Gripper Force Construction Sector Pre- and Post-financial
Crisis. International Journal of Strategic Property
h Hours
Management, 2014.
Kg Kilograms
6. Pilateris, P; McCabe, B. Contractor Financial
LF Lift Force Evaluation Model (CFEM). Canadian Journal of
m Meters Civil Engineering, 2013, Volume 30(3), Pages
487–499.
N Newton
7. Horta, I. M; Camanho, A. S; Da Costa, M.
ROCCO Robot Assembly System for
J. Performance Assessment of Construction
Computer Integrated Construction
Companies: A Study of Factors Promoting
s Seconds Financial Soundness and Innovation in the
SF Sensor Force Industry. International Journal of Production
Economics, 2012, Volume 137, Page 84–93.
TCP Tool Center Point
DOI: 10.13034 / JSST-2015-032 THE JOURNAL OF STUDENT SCIENCE AND TECHNOLOGY 2015 VOL 8 ISSUE 3 33
8. Tsolas, I. E. Modelling Profitability and 11. Gebhart, F; Bock, T. ROCCO—Robot Assembly
Effectiveness of Greek Listed Construction System for Computer Integrated Construction,
Firms: An Integrated DEA and Ratio Analysis. An Overview. European Conf. on Product and
Construction Management and Economics, 2011, Process Modelling in the Building Industry, 1994,
Volume 29, Pages 795–807. Volume 1.
9. Xue, X.; Shen, Q.; Wang, Y.; & Lu, J. Measuring 12. Adeli, H. Smart Structure and Building Automation
the Productivity of the Construction Industry in in 21st Century, in: E.K. Zavadskas, A. Kaklauskas,
China by Using DEA-based Malmquist Productivity M.J. Skibniewski (Eds.), International Symposium
Indices. Journal of Construction Engineering and on Automation in Construction, 2008, Volume 25,
Management, 2008, Volume 134(1), Pages 64–71. Pages 5–10.
10. Pritschow, G; Dalacker, M; Kurz, J; Gaenssle, M. 13. Chamberlain, D; Ala, S; Watson J; Speare, P.
Technological Aspects in the Development Masonry Construction by an Experimental Robot.
of a Mobile Bricklaying Robot. Automation in International. Symposium on Automation and
Construction, 1996, Volume 5, Pages 3-13. Robotics in Construction, 1992, Volume 9.
34 2015 VOL 8 ISSUE 3 LA REVUE POUR LES ÉTUDIANTS EN TECHNOLOGIE ET SCIENCES DOI: 10.13034 / JSST-2015-032