A Survey of Cable-Suspended Parallel Robots and their ...papers.cumincad.org/data/works/att/sigradi2016_484.pdf · Georgia Institute of Technology, USA [email protected]

SIGraDi 2016, XX Congress of the Iberoamerican Society of Digital Graphics 9-11, November, 2016 - Buenos Aires, Argentina

914

A Survey of Cable-Suspended Parallel Robots and their Applications in Architecture and Construction

Fereshteh Shahmiri

Georgia Institute of Technology, USA

[email protected]

Russell Gentry

Georgia Institute of Technology, USA

[email protected]

Abstract

Serial, aerial and solid-linked parallel robots are unable to handle large payloads in building-scale workspaces for on-site

applications and are thus best suited for automated fabrication in plant settings. In contrast, Cable Suspended Parallel Robots

or CSPRs are able to handle large loads and traverse great distances as required on building construction sites. This paper

reviews the existing literature and practice to bridge the gap between our understanding of CSPRs and their applicability to

building-scale tasks such as full-scale concrete printing and building façade installation. The research identifies key activities in

CSPRs fabrication workflows. Using a comparative approach, the paper investigates five CSPR variants and assesses the

performance characteristics. A simple kinematic model of each CSPR is developed and implemented as a Rhino/Grasshopper

script to aid in the performance assessment of each system. The paper concludes with a ranking of CSPR systems and their

likely applicability to full-scale implementation on a construction site.

Keywords: Cable Suspended Parallel Robots; CSPR; Automation; AEC

Introduction

Cable Suspended Parallel Robots or CSPRs are a class of

parallel manipulators in which a mobile end-effector is

connected in parallel to fixed points through actuated cables.

A CSPR consists of three or more motorized winches to

manipulate an object within a defined workspace (Khosravi

and Taghirad, 2014). CSPRs have potential advantages in full-

scale on-site construction as compared to conventional serial,

aerial or solid linked parallel robots. These include the support

for large workspace volumes encompassing entire

construction sites with handling of large loads with low robot

weight and fast process speeds (Lamaury and Gouttefarde,

2013; Sousa et al., 2016). Critical issues for the

implementation of CSPRs on construction sites include

complexities in cable robot kinematics, the trade-off between

redundancy and cable interference (mostly in fully-constrained

systems), the ability to rapidly model and implement ad-hoc

arrangements of winch points and the potential for

environment disturbance in under-constrained CSPRs.

Together these factors make it difficult to determine the optimal

CSPR mechanism for AEC applications (German et al., 2001;

Bosscher and Ebert-Uphoff, 2004). Therefore, despite the

aforementioned advantages, these issue have delayed the

development and implementation of CSPRs and the digital

tools necessary for their implementation in the AEC industry.

The present paper addresses the existing gap between our understanding of CSPRs and key tasks in building construction. To assess common AEC applications, the paper reviews the literature of CSPRs and the key aspects of their mechanisms. The research develops a simplified simulation of five CSPR examples from the literature and provides a ranking of these mechanisms according to their applicability to three

common AEC applications. This initial review paper is the first step in research that will lead to development of a modular CSPR system capable of being implemented in multiple configurations along with an integrated software tool for their programming.

Study Material: Classification

The initial investigation of the academic literature and projects

completed in the area of robotic fabrication indicates that four

key aspects can be used to describe and critique any

fabrication robotic system: (1) robot kinematics, (2) tools and

end effectors, (3) workspace volume and limits and (4)

positioning and motion programming methodologies (Sousa et

al. 2016; Izard et al. 2013; Reinhardt et al., 2016; McGee,

2014). However, as there is no clear classification of these four

categories in the literature, the operation of robotic fabrication

won’t be straightforward for all users in beginning steps and

development phases of projects. Hence, to hand a

straightforward process, this paper put bases on faceted

classification (Afsari, Eastman, 2016) with the purpose of

classifying essential activities existing in such workflows.

Figure 1 depicts the proposed classification for CSPRs. The

aspects are related to the four major aspects as described

above and are further subdivided into sub-classes specific to

CSPRs, to include considerations of: (1) cable stiffness, (2)

cable interference, (3) end-effector payload capacity, (4) multi-

function end-effector (5) workspace reachability, (6)

workspace size, (7) precise positioning and (8) speed range.

Case studies in next section of the paper are evaluated in

accordance with the proposed classification.


915

Figure 1: The classification of essential activities in robotic fabrication

workflow and their subclasses pertain to CSPRs

Background

In order to analyze the performance of CSPRs and evaluate their applicability for any specific application, a general knowledge about concepts such as: 1) cable configuration, 2) degree of constraint, 3) Degree of Freedom (DoF), 4) redundancy, 5) singularity, 6) Inverse kinematics (IK) and Forward Kinematics (FK), is required.

Cable configuration

In first concept, the paper addresses the issue of cable configuration. All CSPRs use a cable coiling mechanism to change the cable lengths for controlling the pose of the end-effector. Cable lengths changes are happened due to the positive forces in under-tension cables imposed on the end-effector. So, cable configuration for a given pose refers to the set of cables that are certainly under tension at the given position and orientation. (Merlet, 2014). This cable configuration in kinematic models is classified based on degree of constraint.

Degree of constraint

Second concept addresses the degree of constraint, CSPRs are classified in two categories: under-constrained and fully-constrained systems. In an under-constrained system, forces like gravity may influence on the end-effector pose. However, in a fully-constrained one the end-effector pose is entirely determined as a function of the cables length. In a fully-constrained model, number of cables is at least one more than DoF. (German, Jablokow, Cannon, 2001; Khosravi, Taghirad, 2014; Moreira, Pinto, Costa, Moreira, March, 2015)

Figure 2: left: fully-constrained – right: under-constrained

Degree of freedom

Third concept clarifies what DoF means. It refers to defined modes in which the end-effector can move. The number of DoF is equal to the total number of independent displacements or aspects of motion. (Rouse, 2009) In CSPRs, end-effector is operated in two or three dimensions and may have translational and rotational DoF. (Bruckmann, Schramm, Mikelsons, Hiller, Brandt, 2008) Although at least n+1 cables are sufficient for n DoF, the number of cables can increase in the system due to the concept of redundancy. (German, Jablokow, Cannon, 2001)

Figure 3: At least n+1 cables for n DoF (German, Jablokow, Cannon, 2001)

Redundancy

Fourth concept explains redundancy. it is a crucial factor which depend on robot application, has to be chosen or avoided in the system. Mathematically, it’s value is equal to the difference between number of cables and DoF. The system is redundant if this difference is more than one. It has to be noted that redundancy is independent of degree of constraint. The redundancy may have negative impacts like increasing the potential of cable interference and collision with end-effector and environment. From the other side, it brings potential advantages for CSPRs such as: 1) the expansion of the workspace 2) increasing the lifting capacity and end-effector speed 3) reducing the load on each individual cable, requiring smaller actuators, 4) increasing the cable stiffness and consequently, decreasing the cables and end-effector vibration, and finally 5) avoiding kinematics singularities. (German, Jablokow, Cannon, 2001; Abedinnasab, Zohoor, Yoon, 2012).

Figure 4: adding one degree of redundancy causes the workspace expansion

Singularity

Based on previous concept, singularity is a key factor that needs to be avoided in CSPRs. In fifth concept the issue of singularity is demonstrated. It refers to a particular

http://www.techtarget.com/contributor/Margaret-Rouse


916

configuration in which the robot gains or loses one or more DoF instantaneously. According to the literature, Jacobian singularity from the kinematical point and force-closure one from a dynamical point of view can occur in workspace. “In general, both categories of singularities involve the force transmission problem, namely, the cable forces fail to balance the wrench exerted on the end-effector due to either the singularity of the Jacobian matrix or the cables’ inability to generate tension at all”. (Diao, Ma, Lu, 2008) As a general note, in a CSPR, singularity mostly arise when the end-effector is near the workspace boundary. (Bruckmann, Schramm, Mikelsons, Hiller, Brandt, 2008). Singularity has a serious impact on calculation of movements, trajectory planning, precision of the system and reaching a given pose. (Zi, Wu, Lin, Zhu, 2012) Accordingly, it is required to consider it as a critical point in solving the inverse kinematics.

Inverse and forward kinematics

Since the key concept in CSPRs is dependency between cables length and end-effector pose, solving inverse and forward kinematics is required in sixth concept. The calculation of cables’ length for reaching a given end-effector pose is named inverse kinematics. Contrarily, forward kinematics is the calculation of the end-effector pose according to the known cables length. (Bruckmann, Schramm, Mikelsons, Hiller, Brandt, 2008) There are kinematic equations to calculate the relation between the end-effector position coordinates and cable lengths. Mostly, in the operation of each application, what users need is end-effector pose rather than knowing the cables length values.

Figure 5: Inverse and Forward Kinematics

Figure 6: IK Solver: Cable Lengths calculation in reaching target position for end-effector

As figure 6 illustrates, in a practical simulation, this study developed a simple 2D fully-constrained redundant CSPR with four cables and four servo motors. The simulation and back-end programming for solving the inverse kinematic is done in grasshopper. User enters the end-effector pose into the application, and IK solver calculates the rotation degree for servo motors, controls the winches and cable lengths to drive the end-effector pose.

Demonstrated concepts provides a key basic knowledge which is required for understanding CSPRs and aims to the investigation of historical case studies in next section.

Historical Examples of CSPRs

Initially CSPR is originated from Stewart platform (Ruiz, Caro, Cardou, Guay, 2015) in which movable platform is connected to a fixed base by six extensible variable-length legs. The lengths are varied to achieve the six DoF motion of the platform (Dasgupta, Mruthyunjaya, 2000). In this study five examples are investigated: 1) NIST SPIDER, 2) FALCON-7, 3) WARP, 4) MACARM, 5) Bosscher Contour Crafting.

NIST Spider

NIST Spider with an octahedral structure is the first under-constrained robocrane able to lift, maneuver and position the end-effector in 6 DoF. It provides a large work volume with avoidance of cables and environment interference. However, fast movement and accurate positioning is problematic due to its degree of constraint and environment disturbance. (Albus, Bostelman, Dagalakis, 1992; Bostelman, Albus, Dagalakis, Jacoff, Gross, 1994; Lytle, Saidi, Bostelman, 2004; Lamaurssssy, Gouttefarde, 2013; Tang, 2014)

Figure 7: The simulation of NIST Spider – Top and front view

FALCON-7

Development of Fast Load CONveyance (FALCON-7) solved existing problem in NIST Spider. FALCON-7, developed for ultra-high-speed assembly applications, with one degree of redundancy created a fully-constrained system in which a vertical rod is located between seven cables. The arrangement of the cables and rod allows obtaining an accurate positioning and a larger motion range using the same motors and pulleys, if longer cables are utilized. In general, there is a direct relationship between speed and vibration in all types of CSPRs. That means the faster speed, the more vibration of the cables and end-effector which causes unstable end-effector positioning and collision with the environment. The proposed cables configuration in FALCON-7 as the first ultra-high-speed CSPR is an effective configuration for suppression of vibration by controlling the intensity of internal forces exerted by cables which in turn improves the structural stiffness of the system. (Kawamura, Choe, Tanaka, Pandian, 1995; Kawamura, Kino, Won, 2000; Tang, 2014)


917

Figure 8: The simulation of FALCON-7 – Top and front view

WARP

Despite existing benefits in the structure of FALCON-7, utilizing a cumbersome rod in system restricts the expansion of the workspace. To remove such a cumbrous element, Wirepuller-Arm Driven Redundant Parallel (WARP) mechanism proposed an optimum cable configuration called T-type in an eight-cable redundant system which is illustrated in figure 8. The intelligence of this system is considering the end-effector geometry as an element which determines cables configuration. Hence, WRAP as an ultra-high-speed system for fast assembly of lightweight objects, provided an open work volume with no demand for cumbrous rod existing in FALCON-7. (Maeda, Tadokoro, Takamori, Hiller, Verhoeven, 1999; Tadokoro, Murao, Hiller, Murata, Kohkawa, Matsushima, 2002)

Figure 9: left: T-Type end-effector (Maeda, Tadokoro, Takamori, Hiller, Verhoeven, 1999) – right: cable configuration in WARP

MACARM

To solve the problem of inaccurate positioning in NIST Spider, another approach was developing a fully-constraint system similar to MACARM mechanism with a cubic work frame that one end of cables is fixed to eight corners and the other ends are connected to the end-effector. (Mayhew, Bachrach, Rymer, Beer, 2005; Beer, Mayhew, Bredfeldt, Bachrach, 2008) Since the end-effector has a low speed range and cables configuration decreases the work volume and increases cable interference, it can’t be an appropriate system in operation of AEC applications. However, the transformation of the MACARM structure to another CSPR like Bosscher Contour Crafting, caused a huge progress in automation of AEC applications.

Figure 10: The simulation of MACARM – front and perspective view

Bosscher Contour Crafting

Bosscher Contour Crafting is developed as a translation-only system for 3D printing. It is developed in two versions and includes twelve active cables. In first version, cables are grouped into four upper cables and eight lower cables. Lower ones are in four pairs of parallel cables. To prevent the cable interference with the building being constructed, the forty-five degree horizontal crossbars are actuated vertically. In the second improved version, those crossbars are fixed to the top of the frame and thus don’t need to be actuated vertically. Instead, the single-cable pulleys are actuated vertically as the construction grows. Although end-effector has just three DoF with limitation in applying to a variety of functions, it has a large work volume and workspace reachability. High translational and rotational stiffness in the system, helps to resist unwanted disturbance and reach accurate positioning of the end-effector. (Bosscher, Williams, Bryson, Castro-Lacouture, 2007; Williams, Xin, Bosscher, 2008)

Figure 11: The simulation of Bosscher Contour Crafting version 1 and 2 – front and perspective view

Given evaluation of case studies verifies that cable configuration has key influence on different aspects of CSPR mechanisms and consequently on their application types and efficient operations. In order to investigate case studies’ performance, a comparative assessment of subclasses, proposed in figure 1, is provided in next section.

Comparative Assessment Using the simulation results to assess CSPRs for three common construction applications

A comparison of studied CSPRs’ performance is demonstrated in figure 12. it depicts the strength or weakness of each of five examples in dealing with issues such as cable stiffness and collision, end-effector payload capacity and functions, workspace size and reachability, precise positioning and speed range. In automation of AEC applications, different factors determine the acceptable ranges for mentioned issues. For instance, in robotic pick and place automation, moving large, small, heavy, or hard-to-handle products causes different measures for workspace size, end-effector payload capacity, accurate positioning, etc. In figure 13, three separate diagrams are created for concrete 3D printing, unitized curtain-wall installation and masonry blocks pick and placement. this figure provides an estimation of acceptable range for eight mentioned subclasses in figure 1, and ranked them from one


918

to five. By overlaying two diagrams in figure 12 and 13, it can be concluded how each CSPR example is applicable to the automation of each suggested AEC task.

Figure 12: Performance indicator for five case studies

Figure 13: the acceptable range for parameters in each AEC task

Conclusion In this paper, a brief review of literature on CSPRs is presented

for non-experts in the field of mechanical engineering and

mainly for architects and building constructors interested in

robotics and automation in architecture and construction. For

general understanding of CSPRs, this study has organized

and demonstrated the most key basic knowledge of this class


919

of robotics in six concepts: 1) cable configuration, 2) degree of

constraint, 3) degree of freedom, 4) redundancy, 5) singularity,

6) inverse and forward kinematics. According to given

knowledge of these concepts, variable parameters required for

the analysis of CSPRs is evaluated in five historical CSPR

examples. Those variable parameters are categorized in eight

parts and in a higher level, this study has classified them in

conformity of four highlighted categories in robotic workflows:

1) kinematics simulation, 2) end-effector setup, 3) workspace

analysis, 4) positioning and motion programming. Mentioned

evaluation of case studies demonstrates their weaknesses and

strengths in 1) cable stiffness, 2) cable interference, 3) end-

effector payload capacity, 4) the multi-functionality of end-

effector 5) workspace reachability, 6) workspace size, 7)

precise positioning and 8) speed range. Eventually, in a

comparative approach, their performance is checked for their

applicability to the automation of three AEC applications: 1)

concrete 3D printing, 2) unitized curtainwall installation, 3)

masonry blocks pick and placement.

A limitation of the research is study on dynamic control of

CSPRs that will be addressed in the future study. This paper

mainly has provided a general knowledge about the theory of

CSPR kinematics which will be led to first, a physical

prototyping of unitized axis CSPR and second, the

development of a software in the context of architecture digital

tools in which existing issues in robotic workflows such as

workspace analysis, end-effector design and tool path

programming, positioning and motion programming via inverse

and forward kinematics will be addressed.

References Abedinnasab, M. H., Zohoor, H., & Yoon, Y. J. (2012). Exploiting

Higher Kinematic Performance-Using a 4-Legged Redundant PM Rather Than Gough-Stewart Platforms. INTECH Open Access Publisher.

Afsari, K., & Eastman. (2016) C. M. A Comparison of Construction Classification Systems Used for Classifying Building Product Models.

Albus, J., Bostelman, R., & Dagalakis, N. (1992). The NIST robocrane.Journal of Robotics System, 10(5).

Beer, R., Mayhew, D., Bredfeldt, C., & Bachrach, B. (2008, October). Technical evaluation of the MACARM: a cable robot for upper limb neurorehabilitation. In 2008 2nd IEEE RAS & EMBS International Conference on Biomedical Robotics and Biomechatronics (pp. 942-947). IEEE.

Bosscher, P., & Ebert-Uphoff, I. (2004, April). A stability measure for underconstrained cable-driven robots. In Robotics and Automation, 2004. Proceedings. ICRA'04. 2004 IEEE International Conference on (Vol. 5, pp. 4943-4949). IEEE.

Bosscher, P., Williams, R. L., Bryson, L. S., & Castro-Lacouture, D. (2007). Cable-suspended robotic contour crafting system. Automation in Construction, 17(1), 45-55.

Bostelman, R., Albus, J., Dagalakis, N., Jacoff, A., & Gross, J. (1994, August). Applications of the NIST RoboCrane. In Proceedings of the 5th International Symposium on Robotics and Manufacturing (pp. 14-18).

Bruckmann, T., Schramm, D., Mikelsons, L., Hiller, M., & Brandt, T. (2008). Wire robots part I: Kinematics, analysis & design. INTECH Open Access Publisher.

Dasgupta, B., & Mruthyunjaya, T. S. (2000). The Stewart platform manipulator: a review. Mechanism and machine theory, 35(1), 15-40.

Diao, X., Ma, O., & Lu, Q. (2008, September). Singularity analysis of planar cable-driven parallel robots. In 2008 IEEE Conference on Robotics, Automation and Mechatronics (pp. 272-277). IEEE.

German, J. J., Jablokow, K. W., & Cannon, D. J. (2001). The cable array robot: Theory and experiment. In Robotics and Automation, 2001. Proceedings 2001 ICRA. IEEE International Conference on (Vol. 3, pp. 2804-2810). IEEE.

Izard, J. B., Gouttefarde, M., Baradat, C., Culla, D., & Sallé, D. (2013). Integration of a parallel cable-driven robot on an existing building

façade. InCable-Driven Parallel Robots (pp. 149-164). Springer

Berlin Heidelberg.

Kawamura, S., Choe, W., Tanaka, S., & Pandian, S. R. (1995, May). Development of an ultrahigh speed robot FALCON using wire drive system. In Robotics and Automation, 1995. Proceedings., 1995 IEEE International Conference on (Vol. 1, pp. 215-220). IEEE.

Kawamura, S., Kino, H., & Won, C. (2000). High-speed manipulation by using parallel wire-driven robots. Robotica, 18(1), 13-21.

Khosravi, M. A., & Taghirad, H. D. (2014). Robust PID control of fully-constrained cable driven parallel robots. Mechatronics, 24(2), 87-97.

Lamaury, J., & Gouttefarde, M. (2013). A tension distribution method with improved computational efficiency. In Cable-driven parallel robots (pp. 71-85). Springer Berlin Heidelberg.

Lytle, A. M., Saidi, K. S., Bostelman, R. V., Stone, W. C., & Scott, N. A. (2004). Adapting a teleoperated device for autonomous control using three-dimensional positioning sensors: experiences with the NIST RoboCrane.Automation in Construction, 13(1), 101-118.

McGee, W. (2014). Robotic Fabrication in Architecture, Art and Design

2014. M. P. de Leon (Ed.). Springer.

Maeda, K., Tadokoro, S., Takamori, T., Hiller, M., & Verhoeven, R. (1999). On design of a redundant wire-driven parallel robot WARP manipulator. InRobotics and Automation, 1999. Proceedings. 1999 IEEE International Conference on (Vol. 2, pp. 895-900). IEEE.

Mayhew, D., Bachrach, B., Rymer, W. Z., & Beer, R. F. (2005, June). Development of the MACARM-a novel cable robot for upper limb neurorehabilitation. In 9th International Conference on Rehabilitation Robotics, 2005. ICORR 2005. (pp. 299-302). IEEE.

Merlet, J. P. (2014, May). Checking the cable configuration of cable-driven parallel robots on a trajectory. In 2014 IEEE International Conference on Robotics and Automation (ICRA) (pp. 1586-1591). IEEE.

Moreira, E., Pinto, A. M., Costa, P., Moreira, A. P., Veiga, G., Lima, J., ... & Costa, P. (2015, March). Cable robot for non-standard architecture and construction: a dynamic positioning system. In Industrial Technology (ICIT), 2015 IEEE International Conference on (pp. 3184-3189). IEEE.

Reinhardt, D., Saunders, R., & Burry, J. (2016). Robotic Fabrication in

Architecture, Art and Design 2016. Springer International Publishing.

Ruiz, A. L. C., Caro, S., Cardou, P., & Guay, F. (2015). Arachnis: Analysis of robots actuated by cables with handy and neat interface software. In Cable-Driven Parallel Robots (pp. 293-305). Springer International Publishing.

Rouse, M. (2009). Degree of freedom. http://whatis.techtarget.com/definition/degrees-of-freedom

http://whatis.techtarget.com/definition/degrees-of-freedom


920

Sousa, J. P., Palop, C. G., Moreira, E., Pinto, A. M., Lima, J., Costa, P., ... & Moreira, A. P. (2016). The SPIDERobot: A Cable-Robot System for On-site Construction in Architecture. In Robotic Fabrication in Architecture, Art and Design 2016 (pp. 230-239). Springer International Publishing.

Tadokoro, S., Murao, Y., Hiller, M., Murata, R., Kohkawa, H., & Matsushima, T. (2002). A motion base with 6-DOF by parallel cable drive architecture.IEEE/ASME transactions on mechatronics, 7(2), 115-123.

Tang, X. (2014). An overview of the development for cable-driven parallel manipulator. Advances in Mechanical Engineering, 6, 823028.

Williams, R. L., Xin, M., & Bosscher, P. (2008, January). Contour-crafting-cartesian-cable robot system concepts: workspace and stiffness comparisons. In ASME 2008 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (pp. 31-38). American Society of Mechanical Engineers.

Zi, B., Wu, X., Lin, J., & Zhu, Z. (2012). Inverse kinematics and singularity analysis for a 3-DOF hybrid-driven cable-suspended parallel robot. International journal of advanced robotic systems, 9.

Documents

A Survey of Cable-Suspended Parallel Robots and their ...papers.cumincad.org/data/works/att/sigradi2016_484.pdf · Georgia Institute of Technology, USA [email protected]