Proceedings of IMECE08 2008 ASME International Mechanical Engineering Congress and Exposition Proceedings of IMECE2008 October 31 - November 6, 2008, Boston, Massachusetts USA 2008 ASME International Mechanical Engineering Congress and Exposition October 31-November 6, 2008, Boston, Massachusetts, USA

IMECE2008-67003 THE ULTIMATE EXPERIENCE IN LEARNING ROBOTICS: BUILDING ROBOTS IN A ROBOTICS COURSE Nael Barakat, Ph. D., P. Eng. Assistant Professor School of Engineering Grand Valley State University 301 W Fulton St. – KEN 343 Grand Rapids, MI. 49504 Voice: 616.331.6825 Email: [email protected]

ABSTRACT Most engineering schools currently include a curriculum component that introduces students to the field of robotics. Multiple methods and techniques are used by engineering educators to help students gain familiarity and interest in robotic systems and their applications. However, very rarely the students get the opportunity to gain the ultimate experience of applying acquired knowledge of the field through building an actual robot. This is because building a robot during a college course involves multiple challenges including robotic systems high complexity and the requirement of combining multiple knowledge bases. Students studying robotics end up, at the most, programming purchased robots, or simulating robots using software, but not actually going through the realities and challenges of putting the system together and making it functional to the point of experimenting with it. In this paper, a unique experience in learning robotic systems and building actual robots is presented. This experience is made available in an elective course on robotic systems engineering at Grand Valley State University (GVSU), School of Engineering (SOE). The produced robots are two or three jointed arm configuration robots, controlled by a programmable microcontroller and built based on classroom gained knowledge. In the classroom, the students learn the kinematics and simplified dynamics of robots, as well as other related topics. In the laboratory, the students are required to apply the learned concepts of kinematics and design in combination with control systems to build a robot that will help them understand and demonstrate these

concepts. The course final projects include robotic systems that are built or integrated by teams of students. These projects provide a range of challenges that extends from mechanical design to control systems. The projects are taken up by teams of students having diversified interests and skill bases within the course. The final outcomes of the course are working robotic systems that can demonstrate the students’ knowledge and interest, which the students use significantly as a proof of their competence level when putting together their resumes to move into the next level of their careers. From an educational angle, the course provides the students with an opportunity to combine multiple knowledge sets, skills, and interest to gain the ultimate experience in education: producing a functional system to specifications. KEYWORDS Robotics, Building robots, Teaching robotics. INTRODUCTION The field of robotics signifies a remarkable outcome of engineering ingenuity. Robotics engineering requires a combined knowledge of multiple engineering disciplines and a good grasp of how engineering systems work coherently to achieve a complex goal or multiple objectives. Robots are complex engineering systems that provide a range flexibility to be utilized in an endless array of applications [1]. Robotics is a term used to indicate a field that deals with robots, in all its forms, in addition to peripherals and joining devices to make up a complete functional system [2]. Robots have been around for a while and have been advancing at a Copyright © 2008 by ASME 1

Copyright © 2008 by ASME

challenges of putting an engineering system together and making it work is an invaluable experience with a significant educational value. During this course, students start by building simple robots to experiment with them and apply the theoretical concepts from the lecture material. The theoretical part runs at a faster pace than the lab part and finishes earlier than the lab part. This helps the students gain the needed background and put it into practice just in time. Cost effectiveness and real life experience starts at this stage where the design of the robots is constrained by the available components in the lab storage. The students are required to pick components from the lab storage and tailor their systems around them. The final robots are stationary serial manipulators with two or three links, the relevant joints, and the required actuators. A controller is required to run the robots and some sensors are needed to provide feedback for these controllers. Engineering concepts gained in the class room are applied to these robots intermittently. These concepts include forward and inverse kinematic modeling, trajectory planning, and simulation of robots using the models. The applications also include the production of control algorithm, relevant to the chosen controller, to achieve simple tracking motions of suggested trajectories. At a later part of the course, students move on to form teams and engage in projects that deal with complete robotic systems. The final outcomes of these projects are functional systems that are used by an external or internal sponsor. Besides demonstrating the gained knowledge, the projects provide the student with opportunities to undertake the parts which interest them individually while gaining the ultimate experience of building an actual system with a team of engineers. Most of these activities were carried out using very tight or almost no budget and utilizing components that were either donated or available in the school storage rooms.

speed compatible with the advancement of computer microprocessors and computational power. Therefore, robots have become a vital industrial element and have found countless other applications [3]. Today, robotics can be found in almost all engineering schools. Engineering curricula incorporate robots as part of engineering education and applications for many reasons. Among these reasons is that robotics is an excellent application of multi-disciplinary engineering systems. Other reasons include the wide range of robotics applications and popularity. Nevertheless, the level of emphasis and the depth of knowledge in the field of robotics can vary significantly from one school to the other, depending on many factors. The mission of the school, available expertise, funding, room in the curriculum, and support, are some of these factors. From these factors stems all the bases of the constraints surrounding robotics related courses. In addition, one of the eminent challenges in teaching robotics is always present in the lab and hands-on parts. Providing a fulfilling opportunity for the students to apply engineering knowledge gained in multiple courses, and experience building an engineering system with a full educational value, while under these tight constraints, is quite tricky of a task. The large funding and support required to carry out these activities and the limitation in time imposed by packed engineering curricula and limited course time are only some of the hurdles to deal with, before a meaningful educational experience in robotics can be provided. However, these challenges did not stop creative engineering educators from producing techniques and methods that help educate students in the field of robotics while providing them with a practical experience in an intriguing and challenging atmosphere. Some teaching techniques are based on software simulations. Others utilize available robotic systems in the market to train students. Design and build of simple robots has also been reported as a successful technique in robotics education [4]. In this paper, a course model that combines theoretical robotics knowledge to the ultimate experience of building robots, while under most of the abovementioned constraints is reported. This course was proposed and run at the School of Engineering (SOE) of Grand Valley State University (GVSU). The paper includes an explanation of the course model, the course content, the students’ activities, and samples of the course results. A summary of the gained experience and remarks on opportunities for improvement are provided at the end of the paper.

COURSE MECHANICS and CONTENT The experience reported in this paper is included in a course entitled: “Robotic Systems Engineering.” The course was offered as an elective at the School of Engineering (SOE) of Grand Valley State University (GVSU), during the summer of 2007. The course is also running this year (2008) during the summer semester. The course consists of 3 credit hours (3 hours/week) in-class lecture/discussion and one credit hour (3 hours/week) lab. The students enrolled in this course were mainly senior Mechanical Engineering (ME) students as well as Product Design and Manufacturing (PDM) Engineering students. Other engineering students are welcome with permission of the instructor. The course objectives are for the students to: 1. Learn the terminology, definitions, main components, safety issues, and types of robotic systems as well as their industrial applications. 2. Implement spatial kinematics to model and analyze different robots and extend the method for use in general machinery modeling and analysis.

COURSE MODEL At the SOE of GVSU, an elective senior level course to teach robotics engineering was proposed and run for the first time during the spring and summer semesters of 2007. The course model features cost-effectiveness combined with hands-on experience. This model is based on a combination of traditional techniques coupled with a focus on providing the students with the ultimate experience of designing and building real-time robots. Taking the students through all the

2

Copyright © 2008 by ASME

Copyright © 2008 by ASME

experiences while allowing them to choose the parts that match their interests and excel in it.

3.

Analyze robot kinematics to define their specifications and control their trajectories. 4. Design and layout an automation workcell, select, and integrate appropriate components for it, including a robot. 5. Learn the characteristics and basic principles of different types of sensors and actuators used in conjunction with robotic systems. 6. Implement the learned knowledge in building a robotic system and testing it. A topical outline of the course is listed in table 1. Table 1: Topical outline of the robotic engineering course, at GVSU-SOE.

NO.

TOPIC

1

Introduction, history, definitions, safety, terminology, fundamentals, spatial description and transformations, programming basics.

2

Forward and inverse transformations, forward and inverse kinematic modeling.

3

Flexible automation systems design and simulation using robotics.

4

Robot design and selection.

5

Trajectory planning.

6

Differential motion, Jacobians, and velocity.

7

Actuators, sensors, basic automation elements, and control architecture.

The prerequisites of the course are dynamic systems modeling and control. As presented in table 1, the theoretical part of the course uses open chain manipulators as a base for studying kinematics and moving into general control and trajectory planning. The lab experience was divided into three parts. The first part included a limited amount of guided tutorials and open ended experiments using industrial robots. The objective of this part was for the students to gain familiarity with robots’ programming and application using more than one platform as well as give them a chance to go through a part of the theoretical side allowing them to start applying it. The second part included the building and testing of simple robots. The objective was to provide the students with a hands-on opportunity to apply the knowledge gained in the course and experience the challenges of an engineering system. The third part was to execute a robotic system project on an industrial that involves a team of students. The objective of this part was to expose the students to real system integration

3

COURSE RESULTS In the summer of 2007 and during the first run of the course, the lab part started with one tutorial to utilize and program the only industrial robotic manipulator that was available at that time. Following the tutorial, the students were requested to solve an actual industrial problem using the robot and simulate the process using generic software. In the second run, two new robots were added to the lab. One of the robots was purchased and the other was a refurbished old industrial robot which was the outcome of a successful project from the previous year. Moreover, commercial simulation software is now incorporated for the students to utilize using a guided tutorial and an open ended industrial problem. The second part of the lab included building a twojoints robot using material available at the school and controlling its motion using microcontroller knowledge gained in a previous course [5]. An additional constraint was imposed which required that the joints in the robot had to be different in type. Once these robots were built successfully, the next task was to analyze them and apply forward and inverse kinematics modeling to them. Simulation of the models was required for verification and comparison. Students had the choice of simulation software and most of them used MATlab. The robots had actuators that were gathered from the shelves of the control systems lab. The controller used was mainly an Atmel ATMega microcontroller included in a thumb board that the students use in a previous control course [5]. The choice of controller was left to the students and one team chose to use a different controller that matched the components they found. These were later used to apply trajectory planning and general motion control strategy. The students were very creative when forced to use only available equipment and parts found in the school labs. Figure 1 represents an upright configuration robot as one of the robotic manipulators built by students in team 1. The manipulator has one rotational joint followed by one translational joint, both actuated by DC motors. The controller used for this manipulator was the Atmel ATMega microcontroller. The work envelope of this robot is a thin walled cylinder. Figure 2 presents another robot built by team 2 consisting of one rotational joint followed by a translational joint and a two-finger gripper. The manipulator configuration was called the hanging robot with a frame carrying it upside down in a gantry arrangement and the work envelope of it is shaped like a disc. The translational joint is an old power screw with a DC motor. Figure 3 shows a different robot that was built based on the equipment found in the lab storage. The robot has a gantry configuration with the first joint in the robot is a translational joint based on a power screw and a linear precision track combination actuated by a stepper motor. The driver and controller of the motor were found in the lab and the relevant software was installed on a PC allowing motion profile programming of the motor. For the Copyright © 2008 by ASME

Copyright © 2008 by ASME

students were very proud of it to the point of publishing movies of it on the internet. The third project focused on system integration. A

second joint, the team decided to use a rotational joint actuated by another stepper motor because of the extra capabilities that could be exploited in the driver and controller of the first motor. The robot is built in an upside down arrangement and the work envelope for it is a planar ellipse. A fourth team accepted skipping this part of the lab and going to directly to the major project because their project was to build a 5-axis robotic manipulator for demonstration purposes. All these devices and material was either made or found by the students in the machine shop and the lab storage.

Prismatic joint

Rotational joint

PROJECT EXPERIENCE The last part of the lab was about building a robotic system and demonstrating its functionality. The teams formed for the second part of the lab decided to keep the same formation. The list of projects executed by the teams is shown in table 2. Table 2: List of course projects executed by the students in the robotics course.

No. 1 2 3 4

Project title Flying probe Sparky II Pop vending cell Material sorting cell

Figure 1: Upright robot built by team 1 in during the lab.

The first project was sponsored by the SOE – Department of Electrical Engineering. A mechanical platform was needed for building a flying probe that would enable the automated testing of circuit boards and allow future expansions. The project had more emphasis on the mechanism design part and less on the control system. The challenge was in maintaining the required precision of positioning of the probe, which is at 0.5 mm. The team was successful in building the platform and demonstrating its functionality using modular stepper motors that can be replaced easily. A patent application for the device has been pursued and therefore no pictures are presented for it in this paper. The second project was to build a 5-axis robotic manipulator that would be used for demonstration and education. The project was sponsored by the Michigan Society of Professional Engineers (MSPE) and the funds were matched in-kind by the GVSU-SOE. A picture of the robot is shown as figure 4. The robot has two AC motors with Variable Frequency Drives and three DC motor with a twofinger gripper. The controller of the robot is an Allan Bradley PLC. A remote control stand is included to guarantee safety of the user, especially when used by middle and high school students. The robot is currently controllable from the stand and can run automatically to demonstrate some motion sequences, including programmed pick and place actions. The robot has also the capability of expanding and following complex trajectories. This robot was built from scratch and the project involved all aspects of robotics knowledge. The

Rotational joint

Prismatic joint

Figure 2 Hanging robot built by team 2.

donated ASEA IRB-G6 robot was inspected in addition to an existing ASEA IRB-L6 that was in the lab sitting in a corner. The students produced one working robot and a control cabinet of the two and built a cell around it. The cell included a donated rotary table and donated light curtains. Control of cell was achieved through a PLC and an HMI. The final cell would allow a user to select a pop can from a menu and the robot would pick it from a tray, drop it on one end of the rotary table, the table would rotate and make it available to the

4

Copyright © 2008 by ASME

Copyright © 2008 by ASME

user through a window guarded by a light curtain.

This

inches long pipes. The sorting would be between plastic and RELAY FOR DIRECTION CHANGE

ULTRA 5000 CONTROLLER 24 VDC FAN

SERVO MOTOR & POWER SCREW

SOLENOID VALVE FOR GRIPPER

VARIABLE POWER SUPPLY

Prismatic joint ULTRA SONIC SENSOR

Rotational joint 12-24V MOTOR

POTENTIOMETER GRIPPER

Figure 4 Gantry robot built by team 3. metal pipes. The student built a ramp to help present the parts to the robot, which will pick the pipe and place it on a traveling tray over the conveyor belt. At a certain point on the belt, sensors were placed to detect the existence of a part and its type then the part would be picked by the gantry machine and placed in a basket based on its material. The cell was controlled using a PLC unit. Figures 7 and 8 represent pictures of this cell.

Sparky II robotic manipulator

Steel mobile table

PROJECTS’ COMPARISON Each project that was executed by a team of students was unique in nature. The first project included challenges in design of mechanical system for precision motion. The robotics course knowledge came in handy in this project where the students applied it in defining the motion characteristics of the machine and the links/joints interaction. The second project included major challenges in designing a system of multiple links and joints and controlling its motion to an acceptable level of accuracy and repeatability. This is in addition of making the system durable and reliable. Building a complete 5-axis robotic arm from scratch by these students was not an easy task. The third project difference was that it challenged the students in the area of system integration including systems from different technology bases and from different generations. The challenges included planning of the cell and motion sequence as well as controlling the complete cell using

Remote control stand

Figure 3: Sparky II demo robot produced in project 2 during the robotics course.

project emphasized system integration and was the attraction of the end-of-semester projects demonstration held annually at the SOE of GVSU. Pictures of the workcell are shown in figures 5 and 6. The fourth project was also a system integration project. A Mitsubishi robot was integrated with a donated gantry robot and a conveyor belt to form a cell for sorting 3

5

Copyright © 2008 by ASME

Copyright © 2008 by ASME

course mechanics and organization, course content and books/literature, and instructor role. However, the bigger portion asks question requesting the students to describe their

a main control unit to achieve the desired goal. This was also the summary of challenges in the forth project. The projects were similar in the following aspects: 1. All projects had to have a clearly identified real industrial problem related to robotics or automation. 2. All projects had to solve the problem using actual industrial equipment that relate to robotic systems and automation. 3. Externally funded projects had to go through a mutually approved acceptance criteria with the customer as well as stricter time limits. 4. A working final outcome was the common and first requirement among all projects. 5. All teams had to have pre-identified timelines and milestones. Among the milestones were two progress reports and two design reviews.

Rotary table

SUMMARY and FUTURE DIRECTIONS A robotics course at GVSU-SOE was designed and delivered to combine traditional teaching methods with a hands-on and cost effective approach. The lab part was divided into parts where the students were provided the opportunity to gain the ultimate experience of building robots. Students were requested to fill an end of semester evaluation of the course experience. The questions were designed to cover all course aspects including ABET related issues. These questions include a smaller portion covering

ASEA IRB G6

Figure 6: Pop vending robotic cell.

experience and learning process. It solicits input on what worked and what didn’t as well as input on what they wish to have had that would have made the experience even better and would have facilitated their learning process. These students are in their senior year and such questions allow them to measure and compare their course experience in light of their coop experience in a real industrial environment. As common to all students’ evaluation of courses, the answers carry

Light Curtains

Pop tray

Robot control cabinet HMI and PLC cabinet

Figure 5: A second view of the pop vending robotic cell.

6

Copyright © 2008 by ASME

Copyright © 2008 by ASME

Controller box common themes that the instructor can capitalize on. The students’ feedback at the end-of-semester evaluation was significantly positive. Most of the students kept documentation of their projects to use as part of their portfolio when looking for a job. The lessons learned in this course were directly applicable to real life engineering situations. Design with multiple constraints and working around existing equipment are some of the realities of real world engineering practice. Capitalizing on opportunities for improvement that were obtained in the first course offering, some parts of the course were modified in the second offering. More equipment and tutorials were added to the lab allowing the students to experience multiple platforms and applications. In addition, dedicated robotics simulation software was obtained and provided in the lab. Students were required to simulate a real industrial material handling operation. For this year, it was a machine tending process using a robotic arm tending to machines with different operating times. Another aspect that needed adjustment was the timing between the theoretical part and the modeling and building of robots. Also, projects were front loaded in the course so students can get an early start on the design and selection part. The second offering of the course is running as this paper is being written. The hope is to get more useful feedback to take this course to the next level.

Mitsubishi robot and feeder (at the back)

Gantry robot

Figure 7: Front view of the material sorting cell.

REFERENCES 1. Rehg, James A., “Introduction to Robotics in CIM Systems,” Fifth Edition, Prentice Hall 2. Craig, J.J. “Introduction to Robotics,” Pearson Prentice Hall, Upper Saddle River, NJ. 2005. 3. Nolfi S. and D. Floreano, “Evolutionary Robotics: The Biology, Intelligence, and Technology of Self organizing machines,” MIT pres, 2000. 4. M. Rosenblatt, H. Choset, A. Graveline, and R. Bhargava, “Designing and Implementing a Hands-On Labs for an Introductory Robotics Course: A Case Study in Directed Constructionism,” ASEE Annual conference and Exposition, St. Louis, MO. 2000. 5. Barakat N., and H. Jack, “A Hands-On Approach in Teaching Dynamic Systems Modeling And Control,” IMECE, ASME – WAM, Nov. 2006, Chicago, IL

Figure 8: Rear view of the material sorting cell showing the Mitsubishi robot and the feeding ramp.

7

Copyright © 2008 by ASME

Copyright © 2008 by ASME