Learning physical fields through operating robot ...

LEARNING PHYSICAL FIELDS THROUGH OPERATING ROBOT MOVEMENTS: A CASE STUDY IGOR M. VERNER, IGAL USHIN and EVGENY KORCHNOY Department of Education in Technology and Science, Technion, Israel ABSTRACT This paper considers an electronic development workshop course taught in Israeli high schools (grade I I). The students learn electronic systems in the context of their application to operating a mobile robot. They equip a mobile robot platform with sensor systems, and perform assignments of automatic detection of electrical, sound, and other physical fields. Our educational study examines the effect of learning experiences with different physical environments throughout the course on the students’ understanding of physical field concepts. KEY WORDS: electronics, education, workshop, robot, physical field INTRODUCTION Robots are a category of inechatronic engineering products, capable to perform functions which are normally ascribed to humans, interact with them, und act autonomously in various physical environments. A robotics course at the introductory level of engineering education enables students to acquire a holistic “mechatronic” view of electrical, mechanical and computer engineering. Thus, they become involved in self-directed learning and teamwork. Educational robotics relies on the concept of consfructionbn~ [I], which characterizes learning processes involving a learner in the creation of external and sharable artifacts. Studies show that this approach could be effectively used to educate students of all age and experience, and stimulate their intellectual maturity [2]. The constructionist approach was first applied to teaching robotics in the MIT course [3]. In this paper the authors consider their experiences in teaching robotics as focusing on operating a robot to perform tasks in different physical fields. Our educational study behind this course intends to examine it as a possible approach to learning physical fields through practical activities in robotics media. This practice is expected to reinforce the traditional physics courses, which are divided into separate subjects [4] such as mechanics, electricity, magnetism, optics etc. The pilot robotics course was developed in the Technion Department of Education in Technology and Science. It is included in the 2000 and 2001 school curricula of the Nesher Senior High School for students of the eleventh grade studying a subject “Electronics and Computers”. Electronics and Computers Studies The matriculation certificate on secondary education in Israel includes general subjects (obligatory for all students), and optional subjects to be chosen out of a given cluster.

383

Electronics and Computers is an optional subject given in grades ten, eleven and twelve. It includes the following courses: Switching systems, Electronic development workshop, Microcomputers, and Analogue electronics. Electronic development workshop is intended to support the theoretical courses by a series of experiments with electronic systems. The guidelines of the Ministry of Education specify the topics of the experiments, but leave development of experiments to schools. In many cases the experiments are based on a disciplinary approach without considering either connections or practical applications of electronic systems. Our idea implemented in the course is to study electronic systems in the context of their application to designing behaviors of a mobile robot in different physical environments. The robotics approach has the following advantages: experiments can be made with relevant robot subsystems and include connections between the subsystems; the students are involved in solving problems related to real robotic tasks; practice with robots can increase motivation of the students. In the proposed course the students equip a mobile robot platform with sensor systems, to be used for automatic detection of electrical, sound, and other physical fields. Robot Platform We selected the Rug Warrior Pro [ 5 ] as a platform for electronics experimentation. It is distributed as a kit, including mechanical and electronic component, and a textbook [ 6 ] . More specifically, the kit includes: a MC68HC 1 I A micro-controller, two DC motors, photocells for light sensing, an IR system, a sound detector (microphone), collision detectors (bumper), and shaft encoders. It has two driving wheels and a caster ball to be mounted on a chassis plate. An on-board Liquid Crystal Display (LCD) screen provides information on robot operations in real time. Robot behaviors can be programmed in the Interactive C language. After compilation the programs are downloaded to the robot board via the serial port. Important advantages of the Rug Warrior Pro platform in electronics experiments are: software for robot self-testing, an easy way of attaching external circuits, functions for programming parallel processes, and tutoring supported by the textbook. However, we found the need to improve the kinematic scheme of the robot in order to provide more accurate performance of robot movements (see Fig. I).

Figure I. The adapted Rug Warrior robot

To eliminate generating torques by the back caster ball, we replaced it by forward and back omni-directional wheels. We also significantly reduced robot’s weight by using a suitable battery pack and provided the right position of the robot’s gravity center. A special bearing base was mounted on the robot in order to facilitate integrating external circuits to the board.

384

Course curriculum The course includes experiments in the following topics: switching systems, programmed controllers, analogue electronics, assembling electronic circuits, and energy conversion devices. The conventional course included experiments with switching systems, in which the students implemented logic functions by logic gates. They also analyzed operation of controlled digital devices through both EWB simulation and real modeling by means of a ‘Switching Systems” kit. The students applied assembly language to program basic data processing functions by an 805 1 micro-controller. In analogue electronics the students used EWB simulation software to study operational amplifier circuits such as voltage amplifier, comparator, and summator. Then they assembled the circuits on a matrix board and tested them with an oscilloscope. Practice with energy conversion devices included controlling DC motor speed velocities in open and closed loop modes. In the 2000 school year we developed and implemented an alternative variant of the course, in which part of the experiments were conducted in a robotic context. The students turned to programming a MC68HC1 I A micro-controller of the Rug Warrior to perform different robotic tasks. They studied voltage amplifier circuits with application to a sound detector. They controlled DC motors of robot’s wheels to provide its motion along certain trajectories. Thus the 2000 course consisted in part of general electronic experiments (40 hours), and of a new robotics part (50 hours) detailed in Table I. Table I. Outline of the robotic part of the course Experiments (2 hours each)

Learning topics

1. Mobile robot and i t s components (4 hours)

El. I . Functional testing of robot components i n a self-test mode. El .2. Basic commands for operating robot subsystems.

2. interactive C language and

~ 2 . 1 Programming . o f simple functions E2.2. Integrated robot behaviors.

programming robot behaviors (4 hours)

3. Kinematics o f a robot with two driving wheels (4 hours)

E3.1. Calibrating wheels‘ velocities by shall encoders, and closed loop control programming.

E3.2. Programming or straight and circular movements. Assignment A I (4 hours): Programming ofrobot motion to a given destination position. 4. Sensors and detecting fields (4 hours)

E4.1. Design and building o f photocell, microphone and pyroelectric sensor circuits.

E4.2. Integrating sensor circuits into the robot, testing by a controller. 5 . Robot operating in physical fields (8 hours).

Assignment A2 (6 hours): Finding point sources o f light. sound and heat. and reaching them by a robot.

Topics 1-3 introduce a mobile robot as a mechatronic system. Topics 4-5 deal with sensor measurements and operating robot movements in different fields. In the next section we will consider experiments with a robot and how students study them. Robotics experiments and assignments

In E I . l the students apply a self-test program, which is part of the Rug Warrior‘s software. They learn to download task programs to the robot, and examine robot

385

operations in the self-test mode. The goal of E1.2 is to operate robot motors, sensors and encoders by means of basic control commands. In experiment E2.1 the students program functions to be used in kinematics calculations with focus on practice in Interactive C. The next experiment, E2.2, is in programming behaviors based on concurrent operation of robot subsystems. An example of such a behavior is the task of straightforward motion until obstacle detection by a IR sensor and changing direction of motion to avoid a collision. In the E3.1 experiment the students are assigned to develop a procedure of determining values of velocity control parameters (VCP) for given linear wheel velocities (LWV). They perform this assignment through the following stages: calibration of shaft encoders for measuring LWV; determination of linear wheel velocities for different values of VCP; programming an interpolation procedure of calculating VCP as a function of LWV. In the E3.2 experiment the students programmed the robot to perform two basic movements: linear to a given distance and circular with given radius and angle. This included coordination of motor velocities, shaft encoder measurements, and calculation of a final position and orientation. The next assignment AI required programming a trajectory of the robot to a destination position, as a combination of linear and circular movements. Experiments E4.1 and E4.2 with different sensors were performed through the following general sequence of steps: assembling and testing the sensor circuit, attaching the circuit to the robot and connecting it to the board, measuring field intensities in various robot positions and calibrating the sensor. In assignment A 2 the students had to program t h e robot to find a field source location and reach it. They dealt with different field sources and developed procedures based on three different methods. The first was a method of polar search iterations. Each iteration consisted of the following steps: a) intensity measurements with a 360" rotation of the robot around its center, b) calculating the intensity gradient and estimating the source location, c) linear movement toward an estimated source location. This method was implemented in the procedure of light source finding. The second triangular search method was used to find heat source locations by means of a pyroelectric sensor. It consisted of four steps: a) detecting the direction to the field source (DFS) through a 360" rotation of the robot in its initial location, b) linear movement orthogonal to DFS over a certain distance, c) determining DFS in a new location, d) calculation of the source location in the intersection of the two DFS and liner movement to it. The third method was finding sound sources through an iterative gradient search. At each of the iteration steps the direction of the greatest increase of field intensity is calculated and the robot moves in this direction to a Newton step distance. The gradient vector is determined as follows: - The robot executes linear movements AB and BC in two directions orthogonal to each other (AB IBC). - The directional derivatives of the field intensity function in point B along directions BA and BC are calculated by measuring intensity increments along the segments. These derivatives are the gradient vector coordinates of the field intensity function. Improvements in the 2001 curriculum Based on positive results of the 2000 course, we decided in 2001 to integrate robotics in all its topics. Three principal revisions were made in the 2001 curriculum. First, the

386

general electronic experiments were reformulated in the robotics context and studied in the Sensors and Detecting Fields section of the curriculum (Table I). The new experiments to be studied in the Switching Systems topic referred to the application of switching devices to robot sensor fusion. A new experiment in Analogue Electronics on enhancing the microphone amplifier sensitivity relates to sound detection by the robot. Second revision, the section of robot operating in physical fields was extended by additional concepts and assignments related to robot operation in gravitational and electrostatic fields. We included a tilt sensor building and testing experiment with an assignment of determining level lines on a 3D surface by the robot. A new assignment with regard to an electrostatic field induced on a graphitic paper sheet is performing robot movement along equipotential lines. As a third revision, spatial manipulations were added to the robot assignments of sensor detection and motion. The 2001 course includes an additional part, in which the students practice in assembling and operating computer controlled manipulators from the modular robot kit Rascal [7]. The following topics are considered: - Introducing basic mechanisms and the Rascal kit. Learning activities focus on analysis and synthesis of basic lever linkages and constructing mechanisms equipped with computer-controlled servomotors. Acquisition of construction skills is supported by interactive animations prepared on the Autodesk Inventor software. - Composed mechanisms The students learn kinematics of lever mechanisms and perform project assignments of designing, building and operating mechanisms for drawing different mathematical curves. They are involved in creating mechatronic products through considering various possible solutions, finding effective configurations of the kit’s components and additional parts. Examples of two projects are presented in Fig. 2. The first mechanism is used to draw ellipses (Fig. 2A), the second is for reproducing reverse curves (Fig. 28). B

A

Figure 2 . Drawing mechanisms: (A) ellipsograph. (R)reproducer of reverFe cunes.

Evaluation and assessment In 2000, the course was given in a class of 17 students and assessed through the following formative procedure. In the general electronics part the students submitted lab reports on the experiments. Individual quizzes were also conducted. In the robotic part of the course the students were tested in Interactive C and kinematics. They were required to demonstrate actual performance of the robotic tasks, which they carried out in groups of three, and were asked to present their personal contributions to the group work. Assessment results indicated that the students acquired knowledge and practical experience in electronics consistent with the learning objectives of the Electronic

3 87

Development Workshop. In addition, the students gained important skills related to analyzing, building, and operating mechatronic systems. In the course evaluation, the focus of our educational study was on the following question: To what extent and how can the course contribute to improving students’ understanding concepts of physical fields? We assumed that involvement of students in robotics activities, which integrate experiments with different physical fields, could help them to develop conceptual knowledge in this subject. The connection of practical activities and physical concepts was emphasized throughout the course. For example, in experiments with sensors the students studied characteristics of a field, mechanisms of their conversion to an electric signal by a sensor, and methods of data analysis. Additional lessons on mathematics and physics were given to students with lower prerequisites. Pre-course and post-course tests were conducted to assess the results of learning the subject. The tests comprised qualitative problems on understanding physical concepts (similar to items of the Conceptual Survey in Electricity and Magnetism [4]), as well as tasks on applying methods of quantitative analysis. Test results were as follows: - Most of the students failed or showed low achievements in the pre-course test. - Most of the students succeeded in the post-course test. Their approaches to problems were correct and the mistakes related only to specific steps of the solution. The course evaluation was continued and extended in 2001.

Conclusion This paper proposes a new approach to the Electronic Development Workshop course, which is part of the Electronics and Computer optional matriculation subject in Israeli senior high schools. Accordingly, electronic systems are studied as components of a mobile robot and applied to designing its motion and interaction with different physical environments. In the course experiments and assignments the students equip a mobile robot platform with sensor systems, and program it to perform automatic detection of temperature, electrostatic, and other physical fields. The course has been successfully implemented in the Nesher Senior High School for students of the eleventh grade. Assessment results indicated that the integrated curriculum promoted achieving learning objectives of the electronic development workshop and provided practical experience in mechatronics. We found an important additional effect, namely, experiments with different physical environments throughout the course can improve the students’ understanding of physical field concepts. REFERENCES I. Hnrrl. 1. ;ind

Pcipcrt S. (Ed\.) C ~ ~ f i . s f ~ ~ f ~Ablex ~ f i ~Publishing, ~ ~ ~ i . \ r Norwood, ~i. N I (1991).

2. Verner, I., Ahlgren, D.and Mendelssohn J. “Fire Fighting Robot Competitions and Learning Outcomes: A Quantitative Assessment.” Proc. ASEE Annual Conference, St. Louis (2000). 3. Martin, F. “Building Robots to Learn Design and Engineering. ” Proc. Fronliers in Edtrcafion Conference, Nashville, ‘IN. 12C5, 213-21 7 (1992). 4. Bagno, E., Eylon. B. and Ganiel, U. “From fragmented knowledge to a knowledge structure: Linking the domain of mechanics and electromagnetism.” American J. of Physics, 68(7), SI 6-S26 (2001). 5. URL: http://www.hobbyrobot.com/RugWarrior/ 6. Jones. J., Seiger, B. and Flynn, A. “Mobile Robots, Inspiration to Implementation,” 2nd ed., A. K. Peters. Massachusetts (1999). 7. URL: http://w*ww.robix.com/

388