UpperArm Rehabilitation Robotics System Modeling Using SysML

(Authors)

Aman Jaffari, Cheol-Jung Yoo, Jung-Hoon Shin

(Source)

13(8), 2015.8, 17-27 (11 pages)

Journal of Korean Institute of Information Technology 13(8), 2015.8, 17-27 (11
pages)

(Publisher)

Korean Institute Of Information Technology

URL

http://www.dbpia.co.kr/Article/NODE06502924

APA Style

Aman Jaffari, Cheol-Jung Yoo, Jung-Hoon Shin (2015). UpperArm Rehabilitation

Robotics System Modeling Using SysML. , 13(8), 17-27.

(Accessed)

163.152.133.***
2016/09/17 19:38 (KST)

DBpia ,
.
, ,
.
Copyright Information
The copyright of all works provided by DBpia belongs to the original author(s). Nurimedia is not
responsible for contents of each work. Nor does it guarantee the contents.
You might take civil and criminal liabilities according to copyright and other relevant laws if you publish the
contents without consultation with the original author(s).

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 17
http://dx.doi.org/10.14801/jkiit.2015.13.8.17

UpperArm Rehabilitation Robotics System Modeling Using

SysML
Aman

Jaffari*, Cheol-Jung

Yoo**, and

Jung-Hoon

Shin*

Abstract
Currently, as the application of robots becomes vital in every aspect of the society, the robotics technology will
become dominate soon in the near future. Consequently, robotic systems will increasingly grow in size, number and
complexity, while challenges are rising to significantly model such systems. In this paper, we describes a model
based development approach for robotics system based on System Modeling Language (SysML) and conducts an
empirical case study using an upper-arm rehabilitation robotics system. The proposed approach provide a clear view
of the system architecture and facilitated communication and understanding among the practitioners, which further
simplify the system integration process. Furthermore, it showed an exhaustive utilization of model-based system
engineering methodology and SysML in order to address the challenges within the robotic systems development
domain such as complexity, understandability and reusability.

SysML

SysML

Keywords
robotics, modeling approach, SysML, system engineering, model based system engineering

* Dept. of Software Engineering, Chonbuk National University

** Dept. of Software Engineering, Chonbuk National University,
Researcher of Institue of Culture Convergence Archiving
Received: Dec. 18, 2014
Revised: Aug. 24, 2015,
Accepted: Aug. 27, 2015

Corresponding Author: Cheol-Jung Yoo

Dept. of Software Engineering, Chonbuk National University,
567 Baekjedae-ro, Deokjin-gu, Jeonju-si, Jeonbuk, 54896, Korea
Tel.: +82-63-270-3383, Email: [email protected]

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

18 UpperArm Rehabilitation Robotics System Modeling Using SysML

The rest of the study is structured as follows:

. Introduction

Section

Robotics technology holds the potential to transform

discusses the related works. Section III

describes the modeling approach. Section

presents a

the future of the world and is expected to become

brief discussion about the study. Finally, Section V

ubiquitous over the next decades[1]. It provides

describes the conclusion and future works.

solutions to many of the current and future social

challenges, such as: working in a hazard place or

. Related works

taking care of disabled and elderly people[2]. The

high numbers of aging population and severely

Over the years, a number of different studies have

disabled people is one of the driving forces for the

been proposed for modeling robotic systems. Huckaby

adaptation

development

has introduced a descriptive framework based on

technology[1][2]. As the application of robot becomes

SysML to create an abstracted way of representing

vital in every aspect of human life, the development

robot manipulation tasks at a high level[5]. He also

process gets complex rapidly. Consequently,

has introduced some of the work that has been

of

the

robotics

system

the

demand for modularization and standardization in this

performed

area rises.

existing problem claims about the pitiful state of

To overcome the complexity much effort

has already been putting but the benefits of model

previously.

Herman,

introduces

several

practice in software for robotics[4].

based approach (MBA) specially SysML are not

A Study conducted by[3], presents an overall vision

considered exhaustively[3][4]. SysML is utilized today

of a model-driven engineering approach for robotics

in a wide range of applications[5] and has reached at

that fuses strategies for robustness by design and by

a mature level in many of these domains, but not in

adaptation. According to Abdul Rahman[6], most

robotics. The objective of our study is to define a

robotics systems are complex embedded systems. The

model based approach using SysML in order to:

main idea is the creation of reusable software modules

Utilize the benefits of MBSE & SysML

to allow platform independent design by looking

Shift the focus of robotics software development

things from the system engineering analysis point of

from conventional to model based engineering.

Address the three fundamental challenges

views. Additional contribution to the development of

such as

robotics systems was presented in [7] for component-

complexity, understandability, and reusability in

based robotics engineering. Another study proposes a

robotics domain.

technique and theory that uses creative engineering

The research outcomes are not only determined

robot with UML which applies OOP concept to the

how modularization and standardization of robotic

Graphical User Interface of an engineering robot

systems can significantly deal with the three critical

program[8].

challenges, but also give an awareness of SysML

Up to now, different types of studies have been

suitability for large and complex robotic systems.

proposed for the modeling and development of

To this end, we introduced a SysML based robotics

robotics systems, and

system modeling technique and conducted a modeling

The recent standardization of SysML is suggested as a

case study. The case study is based on an upper-arm

good solution for the robotics community. However,

rehabilitation robotics system, which is controlled by

their purpose is to introduce SysML and to make

the patient using their smartphone for the sake of

comparison with other modeling language. Most of the

convenience.

studies

are

just

few of them utilized SysML.

expressing

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

the

process

with

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 19
disorganized SysML diagrams and then leave the
standardization of the task into one significant model.
This study is different from the previews studies,
since it shows an exhaustive utilization of model-based
system engineering methodology and SysML in order
to address the challenges within the robotics systems
development domain. In addition, we investigated the
suitability of SysML in large and complex robotic
systems development, and then form the standardization of the task into a significant model.
III. Modeling approach

The robot system modeling technique used in this

study is based on SysML and model based system
engineering,

and

follows

the

standard

systems

development life cycle phases[9][10]. It implements a

SysML based robot system modeling beginning with
the requirements diagram that model the system
requirements, block diagram that include also an
internal block diagram of the system structure, and
behavioral modeling such as use case diagram,
state-chart diagram, activity diagram and sequence
diagram. In addition, package diagram is used to
define the system architecture design that allocates the

Fig. 1. Robotics system modeling approach

3.1 A brief outline of the approach

The robotics system modeling technique is based on
SysML and model based system engineering. It
implements a SysML based robotics system modeling,
following the three main phases; requirements analysis,
system functional analysis, and architectural design of
the standard systems development life cycle.

software module into different subsystems.

We attempted to apply all the promising SysML
and model based modeling artifacts for our robotics
system. The subsequent discussion explained about
these artifacts to show how the proposed development
approach fits well in robotic system development. The
proposed development approach is briefly presented in
Fig. 1. The approach is depicted as a single view for
ease of understanding and simplicity. The modeling
approach indicates the two main parts of the study:
the activities in the MBSE and SysML artifacts and
their association.
The subsequent Sections will describe the entire
steps in the proposed approach in detail.

3.1.1 Requirements analysis phase

During the requirements analysis phase, requirements
of the system (business requirements, nonfunctional
requirements, and functional requirements) are gathered,
and organized into requirements diagram to improve
the traceability between the requirements as well as
between other modeling elements.
3.1.2 System functional analysis phase
In the system functional analysis phase, structural
and behavioral models of the system are developed.

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

20 UpperArm Rehabilitation Robotics System Modeling Using SysML

A) Structural modeling

The requirements analysis phase consists of the

Describe the structural part of the system using two

business requirements: what must be delivered to

types of SysML diagrams: block definition diagram,

provide value, functional requirements (what the

and internal block diagram.

system must do) and nonfunctional requirements (how

well it performs). Analyzing the requirements is very

B) Behavioral modeling

significant to the success of any system or software

Describe the dynamic part of the system with the

development activities. Thus, supporting the require-

help of the following four diagrams: use case diagram,

ment modeling is one of the key improvements made

activity diagram, sequence diagram, and state machine

by SysML[11]. SysML allows the representation of

diagram.

requirements as model element to make the system

requirements definition part of the system architecture

3.1.3 Architectural design phase

[12]. The system requirements form the basis for the

subsequent system functional analysis phase.

In this phase, the system architectural design is

In this Section, requirements are gathered and

described, which allocates the software module to a

organized into requirements diagram to improve the

physical architecture. The high-level architectural design

traceability among the requirements as well as the

is defined using SysML package diagram.

other modeling elements. Fig. 2 shows the system

3.2 Requirements analysis

requirements diagram for upper arm rehabilitation

robotics

system,

which

depicts

the

relationships

between requirements and other model elements that

satisfy them.

Fig. 2. System requirement diagram

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 21

3.3 System functional analysis

Fig. 3 shows the block diagram for the upper arm

rehabilitation robotics system. The block diagram is

In the system functional analysis phase, structural

developed based upon the two main characteristics of

and behavioral models of the system are developed.

the system: the Android application and the Robot

The emphasis of the system functional analysis phase

controller. Further, the Robot controller block consists

is the transformation of the identified functional

of three subsystem elements: Ax12 motor used for the

requirements into a coherent description of system

robot joint, Bluetooth module to communicate with the

functions (operational contracts)[10][13]. The SysML

Android application, and Dynamixel_AX library that

artifacts applied in the robotics system functional

control the joint motor. The Android application block

analysis phases are the following: structural diagrams

includes the user interface for interaction between the

such as block diagram and internal block diagram for

user and the system, and the Bluetooth connection that

defining the systems structure, and behavioral diagram

establishes the connection between robot controller and

including

Android operating system.

use

case

diagram,

state-chart

diagram,

activity diagram and sequence diagrams for defining

The Internal Block Diagram (IBD) in SysML is

the systems behaviors. In this section, Fig. 3, Fig. 4,

used to model the internal structure of a block in

and Fig. 7 are referred to some robot modeling

terms of properties and connectors. It emphasize at the

artifacts which are already designed by[14].

detailed structural aspects of the relationship among

the blocks than the block diagram[11][15].

3.3.1 Structural modeling

Fig. 4 shows the internal block diagram that

describes the detailed internal composition of each

The system structure is defined using SysML block

constituent part of the robot system under study. It

diagram and internal block diagram. The block

also expresses the connections between the parts in

diagram describes a system as a collection of parts

detail to indicate how each part fits together in the

that plays certain role in a defined context. The

system design process.

Internal Block Diagram presents interconnection and

interfaces between the parts of a block.

Fig. 3. System block diagram

Fig. 4. System internal block diagram

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

22 UpperArm Rehabilitation Robotics System Modeling Using SysML

3.3.2 Behavioral modeling

control information packet to the robot controller. The

robot controller will receive the packet, if the packet

In this section, the four SysML behavioral diagrams

is the right one and

move the joint to the desired

are used to model the dynamic features of the robot

position, and then send feedback to the Android

system as opposed to the static diagram that described

application.

the static view of the system in the previous section.

A sequence diagram represents behavior in terms of

The four behavioral diagrams are use case diagram,

a sequence of messages exchanged between systems or

activity diagram, sequence diagram and state-chart

between parts of systems[17].

diagram. They are used to present functionality in

terms of how external entities are used the system: to
present flow-based behavior; to present behavior in
terms of a sequence of messages exchanged between
systems or parts of systems; and to present behavior
of an entity in terms of its transitions between states
triggered by events respectively.
The use case diagram is used to convey the use
cases that a system performs and the actors that
invoke and participate in interaction. A use case
diagram is a black-box view of the services that a
system performs in collaboration with its actor[16].
The use case diagram of the robotics system is
shown in Fig. 5. It indicates the use cases, the actor,
the interaction between the user and the use cases and
the relationship among the use cases.

Fig. 5. System use case diagram

The activity diagram is used to specify the system

behavior, with the focus on the flow of control and
the transformation of inputs into outputs through a
sequence of actions. The activity diagram is commonly
used as an analysis tool to understand and express the
desired behavior of a system[16].
The activity diagram for the robot joint move is
shown in Fig. 6 in which, the user first presses the
search button of the Android application to scan
device. If the desired Bluetooth device is found, the
user selects it to connect the application to the robot
controller. If the device is not found, the connection
will not be successful and the user can conduct a
re-scanning for the device. After, the connection is
successful, the user presses the move button from the
application, and then it can send the appropriate

Fig. 6. System activity diagram

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 23
We used the state-chart diagram to describe the state
change upon some internal and external events. These
states are important to analyze and implement the
system accurately and precisely. Fig. 8 illustrates the
system state-chart diagram.

3.4 Architectural design

In this section, we describe the system architecture
design that allocates the software module to a physical
architecture.

To

do

this,

we

consider

different

architectural concepts and allocation strategies based

on the performance and safety requirements that are

Fig. 7. System sequence diagram

captured during the requirements analysis phase. The

basic input for the architectural design is the

The purpose of sequence diagram is to present a

subsystem definition. In the case study, the system

dynamic view of the system in such a way to express

design consists of two subsystems: the Android

a sequence of behaviors and events that occurs over

application and the robot controller as shown in

time. The system sequence diagram is shown in Fig. 7.

Fig. 9.

Fig. 8. System state-chart diagram

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

24 UpperArm Rehabilitation Robotics System Modeling Using SysML

M stands for moderate support, and L means low

support.
Managing complexity is a difficult task and robotics
systems are complex in their nature. The modeling of
the robotics system requirements with SysML diagram
helps to reduce the analysis and design complexity
from the very beginning of the system development
process. Specifically, the complexity is addressed very
well with the block diagram that breaks the large
system into different sub-systems to examine the

Fig. 9. System package diagram

The Android

application

provides a graphical

interface, which receives input data from the user and

sends to the robot controller system. Similarly, the
robot controller can communicate with the motor joint
through the Dynamixel_AX library to deliver the
motion information for execution.

. Discussion

While

conducting

the

modeling

process,

we

observed that MBSE and SysML offer good support

to refine the conceptual design description and
software architecture of a robotic system incrementally.
Furthermore, the design elements represent how the
artifacts collaborate among each other to provide the
product functionalities. This ability of having a clear
visibility of the product architecture facilitates the
team understanding and communication, which in turn
simplify the integration activity. However, our study
shows the benefits of MBSE and SysML in robotic
system development, its application in the area is
limited. Therefore, a major research effort has to be
done in the future in order to enhance the complexity,
understandability and the reusability problems.
Foremost, concern of this study is to address the
three attributes of robotics system, which is considered
as the most crucial in its development process. Table
1 shows the three challenges and SysML support level
in the robotics area. In letter H refers to high support,

internal structure of the system components. Moreover,

the rigorous interactive conceptual development process
based on SysML including domain modeling, use case
development, behavioral and structural modeling
facilitate the decomposition of the system into a
manageable size.
The resulting models are more useful than
traditional documentation because they represent
structure, data and functions along with associated
document in multiple views from different direction.
Beyond the benefits of project documentation, the
SysML models facilitate stakeholder communication,
improve the quality of requirements analysis and
reduce the risk of system design.

Table 1. State the problems and support levels

Challenges
Description
The efficient support of the
complexity, which refers as the
Complexity number and type of relationship
between elements in a robotic
system.
The ability to present and integrate
the system from multiple
Understan- perspectives that increase
dability understanding of the system
across the development team and
other stakeholders.
The capability to support the reuse
of existing models and to design
Reusability models with the intention to be
reusable in the future for other
robotics system development

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

Support
Level
H
M
H

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 25
Understandability

concept

the reusability benefits of MBSE, as well as the

mentioned in our study. SysML and MBSE provide

advantage of using SysML in robot system modeling.

good support of understandability by reducing the

The proposed system modeling approach is beneficial

complexity of the system. The various models that are

for any robotics application domain, since it is highly

developed at different phase of the development

reusable, more comprehensive and less complex.

process can give the participant a better chance to

Moreover, using a general purpose modeling language,

understand the system architecture very well. In

such as SysML is a good choice, due to its consistent

addition,

and traceable diagrams that can highly support model

SysML

is

another

models

are

important

intended

to

be

interpretable by both human and computers. The

reusability.

understandability of the models can be enhanced by

Generally, the study can promote the use and

controlling what and how information appears on the

understanding of both MBSE and SysML as well as

diagrams. Though, the models contain a lot of detail,

provide awareness and encouragement to robotics

but only selected information is relevant to commu-

engineers and practitioners to take a step forward

nicate a particular aspect of the design. Furthermore,

action in using SysML. This can shift their focus

the models are easily understandable even for the non-

from conventional approach to model-based approach

technical users because of their diagrammatical nature.

and motivate them for further study in the area.

Finally, the easy understandability of the system

The future work will focus on extending the

facilitates the reusability of the models. In this respect,

proposed approach to cover the implementation phase

model based system engineering provides the flexibility

and testing phase and applying it with various robotics

and expressiveness to define robotics systems quickly

system case studies to utilize the reusability of the

and effectively, by reusing common entities across

design models. Although SysML has a good support

design projects. Therefore, practitioners can create a

for complexity, understandability and reusability, it is

library of models that can be reusable in future

better to extend SysML profiles to ideally suit to the

project development. This library can be a package

robotics domain. To do this, a deep investigation shall

that contains many reusable subsystems, constraints,

be conducted to investigate the most important SysML

common

types,

features and specifications that well fit to the robotics

dimensions and units. With this, it is possible to

data

types

and

common

value

system development, so that we can have something

reduce the overall cost of the software development

lighter than the SysML, which currently has nine

task compared with the traditional tedious manual

diagrams. This can encourage practitioners to focus

methods. Generally, it can save time and minimize the

only on the important aspect the analysis and design

likelihood of errors or bugs since the reusable models

activities.

are already refined.

References
V. Conclusion and future direction

[1] Georgia Institute of Technology et al., "A

In this study, we have proposed a modeling

Roadmap for U.S. Robotics from Internet to

approach for robotics system and conducted an

Robotics", Robotics Virtual Organization (VO), pp.

empirical case study using an upper arm rehabilitation

1-137, Mar. 2013.

robotics system that is controlled by smartphone. The

[2] SPARC the Partnership for Robotics in Europe,

case study demonstrates the modeling capabilities and

"Strategic Research Agenda For Robotics in

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

26 UpperArm Rehabilitation Robotics System Modeling Using SysML

Europe 2014-2020", iTechnic Ltd and RUR Ltd,

and Software Engineering", IBM Software Group,

pp. 1-108, 2013.

pp. 1-25, Feb. 2011.

[3] C. Schlegel, A. Steck, D. Brugali and A. Knoll,

[14] Seong-Hun Kwak, Sung-Bin Hyun, Hye-Min Noh,

"Design Abstraction and Processes in Robotics:

and Cheol-Jung Yoo, "Design and Implementation

from CodeDriven to ModelDriven engineering",

of Robotic Exoskeleton Controlled by Mobile

Simulation,

Device for Elbow and Shoulder Movements

Modeling,

and

Programming

for

Autonomous Robots, pp. 324-335, 2010.

Rehabilitatio", Journal of KIIT, Vol. 12, No. 3,

[4] Herman Bruyninchx, "Robotics Software: The

Future Should be Open", IEEE Robotic and
Automation Magazine, Vol. 5, pp. 9-11, Mar. 2008.
[5] J. Huckaby and

H. I. Christensen, "A Case for

pp. 15-29, Mar. 2014.

[15] T. Weilkiens, "Systems Engineering with SysML/
UML:

Modeling,

Analysis,

Design",

Morgan

Kaufmann OMG Press, pp. 223-270, 2008.

SysML in Robotics", IEEE International Conference

[16] Lenny Delligatti, "SysML Distailed: A Brief

on Automation Science and Engineering (CASE),

Guide to the Systems Modeling Language",

pp. 333-338, Aug. 2014.

Addinson-Wesley, pp. 1-213, 2013.

[6] M. A. A. Rahman, K. Mayama,

"Mobile Robots
Interfacing,

and T. Takasu,

[17] S. Friedenthal, A. Moore, and R. Steiner,

Control Architectures, Bio-

"Practical Guide to SysML-The System Modeling

Navigation,

Multi

Robot

Motion

Planning and Operator Training", InTech, Chap. 1,

Language, 2nd", Morgan Kaufmann OMG Press,

pp. 1-389, 2012.

pp. 21-38, Dec. 2011.

[7] D. Brugali and P. Scandurra, "Component-Based

Authors

Robotic Engineering (Part )", IEEE Robotics &

Automation Magazine, Vol. 16, pp. 84-96 Dec. 2009.
[8] H. S. Hack, "Creative Engineering Robot Design
with UML", KCI, Vol. 33, pp. 343-349, Oct. 2008

Aman Jaffari
2010 : Associate Degree in
Computer Engineering from

[9] C. Haskins (Editor), "Systems engineering handbook:

Vision University of Jeonju

A guide for system life cycle processes and

2012. 8 : Bachelor of Engineering

activities", INCOSE, Seattle, WA, 2006.

Degree in Computer Science

[10] H. Hoffmann, "Streamlining the Development of

and Engineering from Jeonju

Complex Systems Through Model-Based Systems

Engineering", 31st Digital Avionics Systems
Conference, pp. 1-8, Oct. 2012.
[11] Object management group, "OMG Systems Modeling

University
2012. 9 ~ current : MSc course in Dept. of Software
Engineering, Chonbuk National University
Research interests : model-based system engineering

Language (OMG SysML) Specification", OMG,

Version 1.3, pp. 1-272, June 2012.
[12] M. D. Santos Soares and J. Vrancken, "Requirements
Specification and Modeling through SysML",
Systems, Man and Cybernetics, pp. 1735-1740,
Oct. 2007.
[13] H. P Hoffmann, "Systems Engineering Best
Practices with the Rational Solution for Systems

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)

Journal of KIIT. Vol. 13, No. 8, pp. 17-27, Aug. 31, 2015. pISSN 1598-8619, eISSN 2093-7571 27
Cheol-Jung Yoo
1982 : BSc degree in Computer
Science and Statistics from
Chonbuk National University
1985 : MSc degree in Computer
Science and Statistics from
Chonnam National University
1994 : PhD degree in Computer
Science and Statistics from Chonbuk National
University
2012. 1 ~2013. 7 : Visiting Professor at University of
California Irvine
1997 ~ current : Professor in Dept. of Software
Engineering, Chonbuk National University
Research interests : Software Quality and Testing,
Software Complexity, Software Metrics, CBSD,
Software Development Process, Embedded Software
and Testing, Multimedia Software etc.

Jung-Hoon Shin
1991 : MSc degree in Computer
Science and Statistics from
Chungbuk National University
1999 : PhD degree in Computer
Science from Chungbuk
National University
1997 ~ current : Professor in
Dept. of Software Engineering, Chonbuk National
University
Research interests : Multimedia DBMS, Embedded
System, Software Engineering

| IP: 163.152.133.*** | Accessed 2016/09/17 19:38(KST)