Development of a Learning Simulation Framework for Modular Robot Manipulator

Abstract: According to the growing demand for industrial robotic arms, understanding the basic of industrial robotics is therefore important. This research presents a simulation framework that enables users to apply the theory of industrial robotic arms through practice in simulation, consisting of link, joint, degree of freedom. joint type, coordinate systems, workspace, and industrial robot types by designing the simulation in the form of a modular robot that the users can customize the robotic arm according to the user's needs, whether it is adjusting the size of the link, selecting the type of Joint, viewing the robot's endpoint coordinate system. Pros and cons of different types of industrial robotic arms are also considered.


INTRODUCTION
For learning today, in Thailand, the arrival of Industry 4.0 has led to the use of robots in the industrial system.One of the most commonly used robots is the mechanical arm robot.International Federation of Robotics [1], which collects data on the installation of industrial robots every year, it can be found that industrial robots have been increasingly installed every year since 2011 until now.For Thailand in the year 2011 to 2021, there is a 36% increase in installation from the previous year.With the tendency to use robotic arms more and more in Thailand, several courses have been opened for learning industrial robots, either taught by industrial robot manufacturers themselves or by training centers to support the needs of use.The demands for the use of industrial robots are increasing, causing the researchers to explore various industrial robot training courses.Industrial robotic arms, no matter which company they make, require the same knowledge.
The definition of industrial robots [2] are the machine that can replace humans and is programmable, designed to operate in a variety of industrial environments.These robots are used in industrial processes or as replacements for the older manual processes.This makes it possible to increase the power of industrial production processes, and to perform dangerous processes in place of humans, and increase the efficiency of production processes ( [2], [3]).Industrial robot manufacturers that are deployed around the world, e.g., Yaskawa, ABB, Fanuc, and Nachi.
For those who want to use the industrial robotic arm, the first problem encountered is choosing the good industrial robotic arm to meet the demands.In the absence of basic knowledge, the selection of industrial robots will not meet the needs of the tasks that will be used, causing loss of money without benefit.Therefore, if the users can gain the basic knowledge before choosing a robot, it will make the robot they choose gain the most benefit without wasting too much money or too expensive to invest in industrial robots.
The basic knowledge for those who just want to use industrial robots consists of Payload, Reach Distance and Application.[4] In using Payload, we can know from the work that we will use in the industry, but the Reach or selecting this robot will cover Workspace Position and Orientation.Where the end point of the robot will reach is something to know before selecting an industrial robotic arm.For example, the SCARA robot, which is characterized by speed but has the disadvantage of being able to operate on only one plane, or the articulated robot, which is characterized by an end point, can go in multiple orientations but it is more difficult to control.From the knowledge mentioned above, the use of simulation plays a role in the design, manufacturing process and robot selection, with the advantage that it can reduce trial and error time.The wasted resources can be reduced from obtaining robots that do not meet the requirements.
This research therefore foresees the problem at this point, if there is a framework system that allows students to learn the robotic arms through theory and then can apply it to practice immediately.One of the systems that can apply knowledge immediately is a simulation system that simulates various industrial robots without the real thing, which nowadays is programmed to be used by people to accompany their education.There is RoboDK [5] which collects almost all models of robotic arms to try out, but there are still problems in Configuration, which is limited to those robotic arms causing the learning style to be too fixed.Later on, the program RoboAnalyzer ([6], [7]) can be used up to the level of dynamic analysis, but there is still a limitation that there are not many models to choose.Therefore, it does not respond for learning robotic arms in a variety of ways or without learning challenges.
The simulation system for this research has an important point, which is Flexibility in Configuration.Different types of robotic arm robots that respond to this task are modular robots, which is a highly adaptable robotic system composed of individual, interchangeable modules that can be assembled in various configurations to perform a wide range of tasks.Modular robots can morph and reconfigure their physical structure to suit different applications.This flexibility is a fascinating subject for academic research and experimentation [8] that have advantages in customization, designed to be the system for learning various types of robotic arm.

METHODOLOGY
In order to design the simulation system for teaching, users can understand the basics of industrial robotic arms.Then, users explore the needs of today's industrial robotic arms for using in the simulation.

Manipulator Robot Knowledge for User
For specifying the content used in the design of this simulation system, it must be specified that regardless of the use of different brands of industrial robotic arms, they must be used the same as the content from the survey of the brand's robot training.Various in Thailand have the following content: • Link, Joint, DOF, this activity teaches learners to explain the DOF of the industrial robot through slides and the simulation system.Learners understand the relationship between link joints that cause how many DOFs.• Type of Joint, an activity to show the difference of each type of Joint by letting learners try to control each type of Joint.Learners understand the differences in the types of Prismatic and Revolute Joints and the resulting endpoints.• Coordinate System, Jog, activities to understand the 3D axis coordinate system through adding equipment and controlling the Joint Jog.Learners understand the robot's coordinate system, base, endpoint, and equipment.Linear Joint Twisting Joint • Workspace, adjustment of the size of the Link to see where the robot endpoints will reach.The user understands how the size of the link affects the workspace of each robot.• Type of Industrial Robot, experimental activities of Joint Jog, each type of industrial robotic arm.Learners see the differences of each type of industrial robot, and the relationship between the endpoint and the joint.

Robot Modular Model for Simulation
By designing a modular model for use in simulation, one modular model must be designed to be used in many ways.It is designed by separating the types of joints of the robotic arm that have 2 types; 1) Prismatic Joint, and 2) Revolute Joint [9] Both types can be classified as shown in Figure 1.
From the market survey, it can be found that the types of industrial robots from manufacturers found that among 17 manufacturers, all companies had articulated robots and all 10 companies had SCARA types.In this paper a modular robot is thus designed, that can be assembled into the SCARA robot and articulated robot by analyzing the types of joints of both types of robots for design as follows in Table 1 Based on the structural analysis, joint modules are designed by having input connectors and output connectors connected to each of the 3 modules.
• Linear Joint uses Rack and Pinion drive to change the rotational motion of the motor to sliding (Prismatic joint).• Twisting Joint is designed so that the input connector and output connector are on the same vertical axis.• Rotational Joint is designed with Input connector that is perpendicular to Output connector.
In this paper, the Twisting joint and the Rotational joint are designed so that a single structure can be used with both joint types, different from the position of the input connector.Figures 2  Figure 2: Joint Analysis for SCARA Robot and Articulated Robot.[11] Figure 3: Revolute Joint Modules (developed from the idea of [12]).

Manipulator Robot Knowledge for User
Initially, the program will create a World Coordinate Frame (WCF) at the bottom middle point of the simulation world, with the Z axis pointing up as in Figure 6.
As Later, variables according to Table 2 will be obtained to create a robotic arm with: • Length is the length of the next link.Type of Joint.
• Joint Axis, the axis that allows the Joint to move is compared to the world's WCF in the simulation.• Link Axis, the axis determines where the link will point.
• Joint Type, a variable uses to define the type of Joint.
From Figure 7, an example of creating a 3DOF robotic arm is illustrated.
Figure 8 shows the generation of robot for each point in sequence.
• Link, Joint, DOF in 8, there is a robot status window showing the results of the resulting DOF of the robotic arm according to the number of Link Joints created.• Type of Joint, in the instruction set of the program, the type of Joint will affect the program.In the case of Prismatic Joint, it uses the Translate command in the direction of the Joint Axis and the Revolute Joint uses the Rotate command around the Joint Axis instead.• Coordinate System (Base, End-effector), the end-effector of the robotic arm is obtained by reading the current position of the simulation endpoint, which can be used to calculate the forward kinematic to be used relative to the simulation position.When changing the position of the Base, the end position of the robot will also change with respect to WCF as shown in Figure 9. • Workspace, the simulation system can see the relationship distance between Link and Joint.The size of the Link can be adjusted to see the working range of the robot according to Figure 10.x, y, z, -x, -y, -z Axis for Joint move (Reference to WCF) Link Axis x, y, z, -x, -y, -z Generate Link Axis (Reference to WCF) Joint Type 0,1,2 0=Prismatic Joint, 1=Revolute Joint, 2=Fixed Joint   • Type of Industrial Robot, from Type of Joint, a mechanical arm robot can be built as shown in Table 1.Two types of mechanical arm robot, SCARA Robot and Articulated Robot, as shown in Figure 11.

CONCLUSIONS
This research presents the modular robotic arm simulation program designed from the content for the basis of an industrial robotic arm,   For future work, the simulation system will need to support the modular robotic arm that will be able to display results based on real robotic assembly.It is additionally expected that the simulation system will be developed to provide the knowledge for users who want to produce industrial robots for their own use.

Figure 8 :
Figure 8: Robot Generation for Each Joint in Sequence.
consisting of the relationship between Link Joint, DOF, Joint type, Industrial Robot type.The final result is a simulation program that can create a custom robotic arm according to the needs of the user, showing the difference of each type of robotic arm.This simulation program can give enough knowledge to buy and use industrial robotic arms that are most suitable for work in industrial processes.The advantages and disadvantages of each type of robot are shown in this simulation when selecting the type of the

Figure 10 :
Figure 10: Adjustment of Link Length to Show the Area of Link.

Figure 11 :
Figure 11: Generating Different Types of Industrial Robots.

Table 1 :
Joint Comparison between SCARA Robot and Articulated Robot.

Table 2 :
Input for Generating Robot in Simulation.