ROBOCON 2020 Robot Design Handbook

2.3.2 Try Robot Flowchart UMP BOT Figure 10: (TR) Arduino Program Flowchart 3.0 ROBOT TESTING Table 1 shows the data obtained on determining the amount of pressure (bar) needed for the TR to throw the rugby ball to PR for different distance. Figure 11 shows PR and TR at 6 meters apart. Figure 11: Distance for Passing Ball 195

By taking the average value obtained from the experiment that have been done, the average amount of pressure for the TR robot to throw was calculated and as shown in Table 1. This data is used to estimate how much pressure needed to be pumped inside the air tank. Table 2: Pressure Usage Reading Length be- Pressure recorded in (bar) Average reading tween TR of pressure (bar) and PR (m) Experiment 1 Experiment Experiment 3 5.40 2 5.20 5.37 6 4.93 5 5.50 4.56 4 5.00 4.80 5.00 4.60 4.50 4.60 4.0 DISCUSSION/EVALUATION OF FINDINGS 4.1 UNIQUE DESIGN UMP BOT (a) (b) (c) (d) Figure 12: (a) Front view of kicking mechanism, (b) Side view of kicking mechanism, (c) Front view of kicking mechanism in ready to kick state, (d) Side view of kicking mechanism in ready to kick state 196

Our kicking mechanism applies a bearing as a roller attached to a bicycle pedal UMP BOT gear which is controlled by a motor linked with a chain to pull the kicking foot to a designated angle. A harpoon rubber is also used to add force during the kick. Once the kicking foot reaches the required angle, the bearing will roll through the locking bar of the kicking foot and release the foot. After kicking, the roller is reset to the original position using the same motor in the inverse direction. 5.0 SUSTAINABLE ENGINEERING PRACTICES We prioritize the use of materials as well as sustainable work as best we can in our robots. Most parts and mechanisms from our robots use reusable materials from previous competition. For example, our robot reuses the robots frame from ROBOCON 2019 and its circuit boards were also taken from previous robots. There are also mechanisms in robots that reuses scrap materials available at our university. This practice not only helps in terms of storage of material dumping but this method can also help us in saving the cost of goods and production which is a problem for us. 6.0 CONCLUSION, LIMITATIONS, RECOMMENDATION 6.1 CONCLUSIONS After almost three months of completing the project of both Pass Robot (PR) and Try Robot (TR), we finally accomplished all tasks that were given for ROBOCON 2020. Although sometimes not according to the schedule, we were able to complete the tasks in due time. The objectives and the scope have been achieved based on our research and useful advice given by all the supervisors. While completing this project, there were many problems that we have faced to ensure that the project run smoothly. Fortunately, all of the problems had been resolved by the supervisors, experienced people as well as by the team members. Overcoming these problems has created a strong team spirit within our team. More importantly, we are able to get the best results from what has been identified. For this project, we learned that every theory needs testing for better results. For example, the kicking mechanism that was constructed based on theory alone was not able to reach 6 meters. After several testings and modifications, we were able to make an improvement to our design and the ball finally reaches 6 meters. 197

UMP BOT 6.2 RECOMMENDATIONS From the project, there are many aspects that can be further improved such as: 1. More methodological work is needed on how to robustly capture the impact and outcomes in research. 2. Extended theoretical equations or/and software modeling tools is/are needed to produce better simulations. 6.3 LIMITATIONS Realist evaluation is a relatively new approach and we recognise that there were several limitations to our study. First, we acknowledge that before starting the project we need to finalise items, tools, or hardware to be used. Because of the lack of hardware, we will drag the implementation time and could not accomplish the tasks according to the planned schedule. The other limitations are time constraint and lack of financial assistance in the execution of the project. Acknowledgements This work was conducted under the approval of Student Affairs and Alumni Department (SAffAD), Universiti Malaysia Pahang. The UMPBot team are also particularly grateful to UMPBot advisors, Dr. Asrul Adam, Dr. Ahmad Najmuddin Ibrahim, Dr. Saifudin Razali, Dr. Mohd Razali Daud, Dr. Muhammad Aizzat Zakaria, Mr. Idris Mat Sahat, Dr. Nurul Hidayah Razak, and Dr. Ir. Addie Irawan Hashim for willingness to support the successful team. References 1. ABU Asia-Pacific ROBOCON 2020 Suva, Fuji, (2020), Theme & Rule, Retrieved from https://www.aburobocon2020.com.fj/wp-content/uploads/2019/10/ABU_Robocon _2020_Rulebook_Sep_30.pdf 2. ABU Malaysia ROBOCON 2020 UNITEN, Malaysia, (2020), Theme & Rule, Retrieved from https://roboconmalaysia.com/malaysia-robocon-rules/ 198

199 UMP BOT

UTHM 1 FROM UNIVERSITI TUN HUSSEIN ONN MALAYSIA (BATU PAHAT) 1 Mohammad Afif Ayob, Hazwaj Mohd Poad, Rosley Sawarno, Chia Kim Seng, Tan Zi Hong, Lio Wei Yong, Tan Zhan Feng, Chai Ri Feng, Tan Chong Chiat, Yeong Chern Yuan, Zubair Adil Soomro, Muhammad Iskandar Naziri, Teguh Imamul Azmi Muhkus, Nurulnadhirah Ab Rahman, Nurul Syafiqah Mohd Razali, Haizatul Norsolehah Hairudin Faculty of Electrical and Electronic Engineering, Universiti Tun Hussein Onn Malaysia, 86400 Parit Raja, Batu Pahat, Johor, Malaysia. ABSTRACT This paper describes the development of two mobile robots that have been designed with the abilities to pass, kick, receive, and knockdown rugby balls for the tournament of ROBOCON Malaysia 2020. Initially, SolidWorks was used to design the required mechanisms resulting in a rectangle-shaped Pass Robot with a size of 80 x 65 x 115 cm and a triangle-shaped Try Robot with a dimension of 73 x 84 x 97 cm. Both robots are equipped with omnidirectional wheels by means of mecanum and omni-wheel to easily navigate between obstacles and tackle the given tasks in a short time. Meanwhile, Proteus UTHM 1 8.0 was used to produce the circuit design and the printed circuit board (PCB) layout. Arduino IDE was used to program both robots before downloading the program to the Arduino Mega microcontroller. For the kicking mechanism, a 100 W Vexta brushless motor with the help of a gyro sensor for the aiming angle was employed which resulted in the ball to be consistently passed over the goal post. Based on the friendly match with other team on a duplicated game field, the developed robots and strategy can successfully score an average of 70 points and kick the rugby balls of at least three times in under three minutes. The development of both mobile robots does not compromise the environment due to the usage of recycled materials from previous competitions and recyclable materials from 3D printing. In addition, the developed mobile robots can also be used for training purposes with rugby players. 200

1.0 INTRODUCTION UTHM 1 The purpose of the work presented in this paper is to develop two mobile robots namely Pass Robot (PR) and Try Robot (TR) to compete in the ROBOCON Malaysia 2020 competition. The theme of the competition is a rugby game involving rugby balls to be picked up, passed, received, dropped and kicked. Hence, the mechanisms developed for both robots revolve in the idea of solving such tasks in the shortest amount of time. For example, the PR has the ability to pick up, pass and kick while the TR has the ability to receive and drop the rugby balls. Both PR and TR robots are equipped with mecanum and omni-wheels respectively to achieve omnidirectional movement and for easier control by the operator. 2.0 DETAILED DESIGN 2.1 MECHANICAL DESIGN 2.1.1 PASS ROBOT (PR) The design of the PR in SolidWorks is given in Figure 1. The rectangle-shaped developed from square bar aluminium was used as the robot base. Its size is 60 x 65 x 80 cm before extension and 80 x 65 x 115 cm after extension. The increase in the size of the PR increases the stability whenever the robot passes the rugby ball to the TR due to the pressure from the pneumatic passing mechanism. The air pressure used for this robot is 5 bar/72.52 psi with 0.2 bar decreases per passes. There is a small gate designed from a small pneumatic cylinder acting as a lock to prevent the kicking leg from touching the kick ball before running the kicking program. At the end of the mechanism, a gear system with a ratio of 1:1 is powered by a 100 W BLH5100KC Vexta brushless motor. Four mecanum wheels with individually controlled brushless motors were located inside the base for controlling the movement of the PR. 201

UTHM 1 (a) (b) (c) Figure 1: (a) PR in top view (b) PR in side view (c) PR in isometric view 2.1.2 TRY ROBOT (TR) The TR as shown in Figure 2 uses a triangle base with a total dimension of 73 x 84 x 97 cm with three individually controlled omni-wheels using BLH015K Vexta brushless motors. Three encoders are installed by following the direction of omni-wheel to detect wheel slip as well as to regulate the orientation of the TR immediately while it is moving. The side of TR is fully covered with a 97 cm high net that makes it easier to catch the ball while reducing the momentum of the received ball. Power window motor is used as the ball dropping mechanism. At the front of the TR, there is a small door for the robot to touch the ball down. This door helps in reducing the motion of the falling ball. 202

Figure 2: TR in isometric view UTHM 1 2.2 ELECTRONIC DESIGN 2.2.1 PASS ROBOT (PR) Figure 3 shows the design circuit of the PR. The PR robot is combined with the function of passing and kicking. For the passing motion, the Arduino located in PR needs to receive a signal from controller, before passing, to allow the small DC motor to turn clockwise direction to lift the passing mechanism until the above limit switch is touched. When passing the Try Ball to TR robot, the Arduino will control the pneumatic to switch on. After that, another signal is required from the controller to switch off the pneumatic and the small DC motor will turn anticlockwise to put down the passing mechanism until the bottom limit switch is touched. For the kicking motion, the Arduino located in the PR receives a signal from the controller for kicking motion. Then the PR will move slowly in sideways direction until the IR sensor detect the tee. After that, it will move backward until the IR sensor unable to detect the tee. The PR will then turn 20 degrees and keep moving forward until the tee is in front of the kicking leg. 203

UTHM 1 (a) (b) Figure 3: (a) Schematic diagram and (b) PCB layout of the PR robot 204

When the kicking starts, the kicking leg will turn anticlockwise direction at a slow UTHM 1 speed and the pneumatic will turn on at the same time to close the gate to prevent the kicking leg from touching the Kick Ball. Then, the kicking leg will break after the IR sensor detect the kicking leg at the top position and allow the kicking leg to be pulled down by its weight. After 1.5 seconds stop, the pneumatic will turn off to open the gate and the kicking leg will turn in clockwise direction at a slow speed until it passes the IR sensor at the top position. The kicking leg will continue turn clockwise direction with maximum speed to kick the Kick Ball and only breaks after IR sensor detects it again at the top position. 2.2.2 TRY ROBOT (TR) Figure 4 shows the circuit design of the Try Robot (TR). For the knockdown motion, the Arduino located in the TR received a signal from the controller to place the Try Ball in the Try Spot. Then the Arduino will send output signals to the power window motor to turn in clockwise direction at maximum speed via its digital pin and PWM pin. To return the knockdown mechanism back to its original position, the Arduino need to receive another signal from the controller which will allow the power window motor to move in the anticlockwise direction. (a) 205

UTHM 1 (b) Figure 4: (a) Schematic Diagram of TR (b) PBC Layout of TR To ensure the TR is moving in a straight line and not deviate from the decided path, three rotary encoders were installed in parallel to the direction of each omni wheel. The rotary encoders act as feedbacks for the distance travel by each omni wheel and by adding PID control to the programming. The TR can also auto tune the speed of the omni-wheels to maintain itself on origin path. 3.0 DISCUSSION Our Pass Robot and Try Robot were uniquely designed such that the two robots had been simulated by theory and supported by calculation especially for the kicking leg mechanism of Pass Robot. BLH5100KC 100 W Vexta brushless motor had been used because it is suitable with the track field of ROBOCON 2020. In order to kick the ball which weighs at at 340 g over 1.5 meters height of the goal post at the furthest distance of 8 meters, it needs a 4.42 N total force to be applied on the ball. Force is calculated by the equation F = ma; where ‘m’ is the mass of ball, and ‘a’ is the acceleration of ball [1]. To get the required acceleration of the ball, the projectile motion should be calculated. If the hit point of the ball were assumed to be at 35 degrees with a speed 13 m/s, 16.2 m of total distance is needed to allow the ball to travel in that speed. A maximum height of 2.8 m could be achieved at the half of the total distance travelled which is 8 m. Thus, a weight had been added onto the end of the kicking leg to produce 4.42N of force. 206

BLH5100KC 100Watt Vexta brushless motor could not produce a torque of 8 N/m, UTHM 1 but can provide enough acceleration of 22 ms-2. When a load of 400 g is added to the end of kicking leg and it can produce 8.8 N of force. By using a kicking leg with a length of 55 cm (maximize the radius while not exceeding the dimension required) and force of 8.8 N, a total torque of 4.84 N/m were produced which is slightly more than the minimum requirement from the formula Ʈ = F*r; where ‘F’ is the force, and ‘r’ is the radius of the kicking leg [1]. Besides that, there is also another unique design for the kicking mechanism. The radius of curvature of the concave kicking leg was designed to be 5 cm in order to get the correct kicking angle at the ball. This has a better performance with chain size #25 because it lowers 2.5 times compared to chain size #35. 4.0 SUSTAINABLE ENGINEERING PRACTICES It is important to become aware of the environmental consequences of all mankind’s production activities in order to achieve life quality and to be able to provide favourable environmental conditions for future generations [2]. To ensure that the development of the robots does not compromise the environment or deplete the materials for future generations and it could improve the quality of life for all, both of the robots was developed from reuse existing built assets such as the steel bar that has been used for the model. Most of the materials was used in previous ROBOCON competition, so that it will not be such a waste for environment and in line with the goal which to design the robots for minimum waste. Not just that, the development of the robots also used recycle items such as bottles for tank gas that is used for handling the gas pump for the robots so that it can reduce the pollution. A big breakthrough on the path to a more socially and environmentally responsible, better-regarded global technology industry is making the development of the robots more sustainable. 5.0 POSSIBLE REAL-WORLD APPLICATION Both Pass Robot and Try Robot are useful in order to help rugby players in their training. The epidemic situation of coronavirus pneumonia is hard to predict in the future, hence, in order to help the players to be ready for their next match, using robot to assist in their training would be a better solution. 207

UTHM 1 6.0 CONCLUSION As conclusion, both of the PR and TR robots were successfully designed with the right mechanism and able to perform the tasks by following the competition rules given. The Pass Robot (PR) was designed to pick up the Try Ball and passes the Try Ball to the Try Robot (TR). In addition, PR was also designed as a Kick Robot to kick the Kick Ball into the Goal during the Kicking Step part. For the Try Robot (TR), it was designed to receive the Try Ball in the Receiving Zone and score a Try in one of the 5 Try Spots. Besides, all the team members have contributed in knowledge and idea sharing, skills and time for this project. Each team member played their role in providing mechanical, electrical, electronic, programming and design skills during the production of the two robots and the team gained more skills and knowledge after completing the project. References 1. Saurabh Chauhan, “Motor Torque Calculations for Electric Vehicle”, International Journal of Scientific & Technology Research Volume 4, Issue 08, August 2015, ISSN 2277-8616 2. Fernando Beiriz and Assed Haddad, Environmental Management in Practice, DOI: 10.5772/22500 Edited Volume by Elzbieta Broniewicz, Federal Fluminense University and Federal University of Rio de Janeiro, Brazil, 2011 3. Datasheet: Brushless Motor: Brushless Motors/AC Speed Control Motor, Oriental Motor General Catalog Page 201 208

209 UTHM 1

UTM B UTM B FROM UNIVERSITI TEKNOLOGI MALAYSIA B Shamsudin H. M. Amin, Mohd Ridzuan bin Ahmad, Chong Kar Lok, Chua Shu En, Hassan Mohammad Shaikh Al-Kaff, Hu Jia Choon, Kevin Chin Kai Sian, Leong Kao Hao, Teoh Sun Yi1, Too Yu Rui School of Electrical Engineering, Faculty of Engineering, University Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia. ABSTRACT This report summarises the mechanical, electronic and software design of the two robots, i.e. Pass Robot (PR) and Try Robot (TR) of the University Teknologi Malaysia (UTM) ROBOCON Team B for participating in ROBOCON Malaysia 2020. The design of the robots to carry out each task in the games, e.g. ball picking up, passing and receiving of try ball from PR to TR, scoring try ball in the try spot by TR and kicking the kickball over the conversion post by PR, is discussed. An electronic circuit block diagram and algorithms of the robots are also described. The approaches for overcoming the game challenges and some recommendations which can improve the robots in the future are shared in this report as well. 210

1.0 INTRODUCTION UTM B UTM ROBOCON Team was established in 2002 and our main focus has been on developing robots and actively participating in ROBOCON activities. For ROBOCON 2020, the game consists of three main tasks which are passing, receiving and kicking. Several designs for passing have been tried such as catapult-style passing, flywheel-style passing by inserting the rugby ball between both spinning flywheel and passing by accelerating the rugby ball at a fixed angle. After countless of trials and errors, passing with a fixed angle is chosen by taking the passing range and stability into consideration. It is a hypothetical believed that kicking from Kick Zone 3 is the key to winning, thus simulations of the kicking mechanism using simulation software to determine the angle of kicking, weight and material for the kicking mechanism was performed. A successful ‘receive’ must fulfil three conditions which are the alignment of both robots, constant passing force and distance between both robots. Additionally, the receiving mechanism is equipped with a fail-safe mechanism and can receive the ball successfully even if the conditions are not met to achieve a 100% success rate for receiving. The idea of the fail-safe mechanism is inspired by the mechanism of the unidirectional door. In an attempt to finish the game within the shortest time, a kicking mechanism to shorten the path travelled by the TR is installed. The strategy being adopted is one try followed by one kick and three tries followed by three kicks. This is the fastest strategy for our team to secure four kick balls to win the competition. The first try and first kickball lay a foundation for us to secure the next three kick balls as TR can receive another ball in no time after the kick attempt. The path for TR and PR can be seen in Figure 1. The rules of the games can be obtained elsewhere [1-2]. Figure 1: Path for TR and PR 211

UTM B 2.0 DETAILED DESIGN 2.1 MECHANICAL DESIGN For PR, a four-points omni navigation base is chosen because of the stability as shown in Figure 2. Brushed motors are used for the four-points omni robot. A pneumatic cylinder is used to grip a rugby ball. The power window motor is used to rotate the gripper to the try ball rack to grip the rugby ball. Four rollers wheel are installed on the gripper hand so that the gripper can hold and grip the rugby ball firmly. After gripping the ball, the gripper will rotate 180 degrees to place the rugby ball on the passing platform. Simulation software is used to simulate the passing action to determine the right cylinder size and passing angle as shown in Figure 3. Figure 2: Pass Robot Figure 3: Simulation of Passing Action 212

For TR, a four-points omni navigation base is chosen because of the stability as UTM B shown in Figure 4. Furthermore, the omnidirectional movement is easier to avoid obstacles. Two cylinders are used to actuate the kicking leg to achieve the required kicking force to kick from the Kick Zone 3. A finite element analysis is conducted to perform the stress- strain analysis of the base structure as shown in Figure 5. The receiving mechanism is specifically designed to block the rugby ball from bouncing out by using the spring door hinge attached to the flat bar and the curtain is used as shock absorber and damp shock impulses of the rugby ball. The part that make the receiving cage so unique is that the fail- safe mechanism will prevent the ball from bouncing out from the receiving cage during passing and only allow the ball to go out from receiving mechanism when try mechanism is actuated. After the rugby ball falls into the try mechanism, try mechanism will grip the try ball actuated by cylinder and rotate 180 degrees to put the try ball in the try spot. This try mechanism is used to ensure that the try is valid and follow the rules and the condition where TR must be in contact with the try ball when try ball touches the surfaces of try spot. Figure 4: Try Robot 2.2 ELECTRONIC DESIGN Figure 6 shows a distribution of sensors and architecture of the processing units within the PR and TR robots. In PR, two microcontrollers are used as the mainboard and robot navigation system, respectively. H-bridge power distribution module is connected to the mainboard to control the valve through the valve driver, LED indicators and motors for the rotating part through motor drivers. For navigation, the motors are controlled by the robot navigation system through motor drivers. The motor driver is equipped with an emergency button to enable power cut-off to the navigation when the emergency button is pressed. 213

UTM B Figure 5: Stress-Strain Analysis Figure 6: Electronic block diagram of PR and TR Besides that, a zone selector is used for zone switching in our mainboard so that operator does not need to upload code again when switching from blue zone to red zone and vice versa. Several sensors such as analog sensors, laser sensors, limit switches, encoders and inertial measurements are used in PR to get more accurate positioning. For the laser sensor, current sensors are needed for the mainboard to get the reading from the laser sensors. Limit switches and analog sensors are used to provide feedback to the 214

mainboard while the inertial measurement unit and encoders which consist of an external UTM B encoder and motor encoder are connected to the robot navigation system. The mode selector is used to signal the mainboard so that a specific mode can be selected. LED indicator connected to the mainboard is used to indicate the condition of the robot. To control the PR and TR using a PS4 controller, a PS4 module is designed and communicate with the mainboard. 2.3 SOFTWARE DESIGN Figures 7 and 8 show the block diagram of the TR and PR, respectively. Immediately after the game starts, PR picks up a ball from the try ball rack and passes the ball to TR at the receiving zone. After the first try, TR will move from the try spot to Kick Zone 3 to kick the first ball. After the kicking attempt, TR will move to the receiving zone and receive another ball from PR. After another three successful tries, three kick balls will be requested and TR will perform a goal kick from the Kick Zone 3. Figure 7: Block Diagram of Try Robot 215

UTM B The software design of PR and TR is based on autonomous and manual mode. The operator can switch between the two modes to complete the tasks. In general, the position of the robot is calculated by using two external rotary encoders and laser sensors. In the case of emergency, the operator can stop the robot using the controller. The operator is also able to switch the code from blue zone to red zone and vice versa by using a toggle switch. To finish the game, TR needs to go through ten paths and each path is indicated by LED indicators. Figure 8: Block Diagram of Pass Robot 216

3.0 PRESENTATION OF DATA UTM B A simple result for tuning the PID of navigation of Pass Robot and Try Robot is shown in Figure 9. The PID of the navigation system is fine-tuned until a perfect smooth graph is obtained to prevent any over-shooting of the robot when navigating around the game field. The time taken to navigate from 0 to 3 meters was approximate 3.5 seconds. Figure 9: Graph of Distance vs Time Figure 10: Tuning Interface Figure 10 shows the tuning interface to tune of the robot in a more user-friendly and easier way. The desired parameters are keyed-in and it will be uploaded to the robot by wireless communication using NodeMcu esp8266. It is only for tuning purposes. After all the accurate values are obtained, it will be unmounted from the robot. 217

UTM B 4.0 DISCUSSION OF FINDINGS Our design is unique because we use a cylinder to actuate the kicking mechanism instead of using a motor as we realize that the kicking force of using a cylinder is easier to control and adjust when compared to using a motor. Besides that, we also consider our receiving mechanism as unique since although our receiving mechanism cannot fully absorb the impact of the pass ball, it has a fail-safe mechanism to prevent the ball from bouncing out. We installed our kicking mechanism on the Try Robot and not on the Pass Robot to minimize the path and the time taken. As an example, if we install the kicking mechanism on the Pass Robot, the Pass Robot has to travel a longer path from Kick Zone 3 to the passing zone. On the other hand, installing a kicking mechanism on the Try Robot will reduce the path travelled and time significantly as the Try Robot only needs to move from Kick Zone 3 to the receiving zone to receive the next ball. 5.0 SUSTAINABLE ENGINEERING PRACTICES During the robot designing phase, we reduced the material or metal we use to avoid wastage. Besides, instead of buying new bottles, we collected the bottles from various food stalls to reduce the amount of waste produced. After several testing, there are some prototypes and designs that fail and instead of throwing them away, we reuse the raw material such as aluminium, stainless steel and mild steel. Some of the raw materials will be sent to the factory to recycle as we understand that recycled materials typically require less energy to process compared to developing new materials altogether. For the robot we build, everything regarding electronic and sensors including wires, microcontroller and all the electronic boards are reusable. Thus, we achieve zero wastage for all electronics parts. Some hardware such as motors, coupling, cylinder, tubes are reusable as well. 6.0 CONCLUSIONS To conclude, the development of TR is a great challenge for us as selecting a suitable cylinder for the kicking mechanism requires many trials and errors. With the involvement in ROBOCON Malaysia 2020, we learned a lot of new knowledge in robotics especially about simulation software and force control mechanism using a cylinder. The limitation in the current PR design is that it is hard to synchronize both cylinders that are used for passing which led to unpredictable passing direction. This will increase the difficulty for TR to catch and receive the ball. The recommendation that can be applied to improve the robot is using only one cylinder that is more powerful to eliminate the synchronization problem. 218

Acknowledgements We would like to express our gratitude towards the Universiti Teknologi Malaysia, Faculty of Engineering, especially the School of Electrical, School of Mechanical and School of Computing for giving us support and facilities to develop our robots. Thanks to our team manager, Prof. Madya Ir. Dr. Mohd Ridzuan bin Ahmad for being also our advisor to advise us in terms of technical issues as well as conducting management in our team. Our appreciation also goes to our team members who tirelessly working together developing better robots for the competition. Not to forget, special thanks to our sponsors for supporting us with the preparation for ROBOCON Malaysia 2020. References 1. ABU ROBOCON 2020. (n.d.). Retrieved October 15, 2020, from https:// www.aburobocon2020.com.fj/ 2. ABU ROBOCON 2020 Theme & Rules “Robo Rugby 7s”. Retrieved October 15, 2020, from https://www.aburobocon2020.com.fj/wp-content/uploads/2019/10/ ABU_Robocon_2020_Rulebook_Sep_30.pdf 3. Solidworks. (2020 ver), Dassault Systemes Solidworks Corporation. Accessed: Oct 2020. [Online]. Available: https://www.solidworks.com/sw/ education/6438_ENU_HTML.htm 4. J. Howie (2020, Sep 25). Altium Designer Documentation [Online]. Available: https://www.altium.com/documentation/altium-designer 5. STMicroelectronics Software AB., Sweden. Atollic TrueSTUDIO User Guide, 20th ed. (Jan 2018). Accessed: Nov 2019. [Online]. Available: http://gotland.atollic.com/ resources/manuals/9.0.0/Atollic_TrueSTUDIO_for_STM32_User_Guide.pdf. 219 UTM B

URT URT FROM UNIVERSITI TUNKU ABDUL RAHMAN Kenny Chu Sau Kang, Sim Sheng Wei, Khor Jun Bin, Au Jin Cheng, Ong Chia Koon, Lim Wen Qing Engineering, LKC FES, Jalan Sungai Long, Bandar Sungai Long, 43000 Kajang, Selangor ABSTRACT This report summarises the mechanical, software, electrical and electronic design of the two robots by Team UTAR. This document outlines the approaches taken by the team UTAR to prepare for ROBOCON Malaysia 2020. We are a team with wide variety of engineering background and we aspire to compete in this year’s competition. Our focus is centred around the passing mechanism, gripping mechanism and kicking mechanism of the two robots. Advances in these directions will enable the robot to play the game of rugby with greater efficiency. 1.0 INTRODUCTION The first task of Pass Robot (PR) is to grab a rugby from the ball rack to load the ball into the robot. This action is done by utilising two pneumatic cylinders. After that, the robot will shoot out the ball to pass the ball to Try Robot (TR) by applying two DC motors with roller wheels. For TR, to receive the ball from PR, it is equipped with a funnel to unload the ball into unloading point. When the TR reach the Trial zone, a pneumatic cylinder will push the ball to unload it from the robot to the trial zone. After that, the robot will perform the kicking motion with the aid of a DC motor and spring to generate enough torque to launch the ball over the desired distance. 220

2.0 DETAILED DESIGN URT 2.1 MECHANICAL DESIGNS 2.1.1 PASSING ROBOT The PR is mainly consisting of two main part; 1) the ball gripping mechanism as shown in Figure 1, and 2) the shooting mechanism as shown in Figure 2. Figure 1: Ball Gripping Mechanism This ball gripping mechanism is constructed based on pneumatic actuators to grip the ball and lift it up to load it into the shooting mechanism of the robot. The black colour part of the robot as illustrated in Figure 2 is the moving part that provide friction and grip and hold the ball while it is loaded to the shooting mechanism. Figure 2: Ball Shooting Mechanism 221

URT The shooting mechanism is constructed using two DC motors connected to the roller wheels to provide sufficient torque to project the ball over a certain distance to reach the TR. The platform of the shooting mechanism was fixed to an angle of 30° to provide a suitable projectile motion for the ball to reach the TR. 2.1.2 TRY ROBOT Figure 3 shows the TR which is equipped with a ball unloading part to make the try and kicking mechanism to kick the ball to make a goal. Figure 3: Overview of TR The TR have a funnel made from black PVC net to roll the ball to the unloading point with desired position of the ball. The incline wall of the funnel provides a smooth rolling motion for the ball to load into the unloading point. 222

URT Figure 4: Ball Unloading Part to Make Try The Try motion of the robot is completed by unloading the ball from the robot by applying a pneumatic cylinder to push the ball out from the robot. A layer of PVC net act as a curtain to block the ball from accidentally drop out and ensure the ball, robot and floor were in touch when making a Try. Figure 5: Kicking Mechanism 223

URT The kicking motion is done by extending the spring to a certain length to provide enough torque to kick the ball over a desired distance. A DC motor is used to pull a wire string that is connected to the kick leg to extend the spring. A pneumatic actuator is used to lock the spring when it reaches the desired extension length. Once the actuator is released, the kick leg will be launched by the spring and kick the ball. 3.0 ELECTRONIC DESIGN The electrical and electronic design is split into 4 parts: wheel, shooting, gripper, kicking and touchdown 3.1 WHEEL 12 V DC motors are used as the main drive motor to drive the 4 mecanum wheels on each robot. MD10C motor driver (Figure 6) is used to control the motor powered by providing PWM signals to control the motor speed. The motor (Figure 7) is also equipped with a rotary encoder which operate at 5 V. Therefore, a 5V regulator is needed as shown in Figure 8. Figure 6: MD10C Cytron Motor Driver Figure 7: 12V 170RPM 2.7kgfcm 32mm Planetary DC Geared Motor with Encoder 224

URT Figure 8: LM2596 5V Voltage Regulator 3.2 GRIPPER For the gripper mechanism, there are three pneumatic cylinders. The pneumatic controller relay as shown in Figure 9 has been installed to control the gripping motion to grab the ball. A boost converter (Figure 10) is needed to convert 12 V to 24 V as the pneumatic solenoid is operating at 24 V. Figure 11 shows a 5V DC relay that is used in the gripper design to control the On and Off of the solenoid. Figure 9: Pneumatic Controller Relay Figure 10: XL6009 Boost Converter Figure 11: Relay (5V) 225

URT 3.3 KICKING MOTION For the kicking motion, a high-power DC motor was used in order to generate enough power to pull the spring to a desired extension length. The motor is operating at 12 V and controlled by a MDDS60 motor driver as shown in Figure 12 and Figure 13 respectively. Figure 12: 12V/24V High Power Motor Figure 13: MDDS60 Motor Driver 3.3 TOUCHDOWN MOTION The touchdown motion design is very simple as it only uses a pneumatic to push the ball out of the robot. It used pneumatic relay to control the pneumatic, due to the pneumatic control relay needed a 24 V to trigger it. Thus, a boost converter is used and is connected to 5 V controlled relay which also to control the pneumatic relay. 4.0 SOFTWARE DESIGN Robot Operating System was used as the main platform to control the robot motions and operations. The Robot Operating System (ROS) is a flexible framework for writing robot software. It is a collection of tools, libraries, and conventions that aim to simplify the task of creating complex and robust robot behaviour across a wide variety of robotic platforms. 226

The low level of the robot such as pneumatic cylinders and DC motor are control by URT the Arduino which receive commands from ROS Node. ROS Core is in charge of receiving command from a remote keyboard and move to the desired position. The locomotion and localisation of the robots are done based on the coordinate of x and y axis that assigned for each function and motion. This enable the robot to operate autonomously according to the game plane that set in the script. The coordinate position of the robot is determined by the count of incremental rotary encoder that attached to the motor. By converting the distance per revolution of the motor, the distance travel and current position of the robot can be determined. 5.0 TESTING AND RESULTS From the test run that had been done, both robots are able to perform the tasks correctly with small margin of error. All the tasks from passing the ball, making a try and kick the ball can be carried out by the two robots. The robots are able finish the challenge within 3 minutes. The TR is also able to score the kick from Zone 3 with high consistency and accuracy. 6.0 CONCLUSION AND LIMITATION In short, the robots are able to carry out the tasks correctly with proper control. The robots designed are within the rules and regulation fixed by organiser. However, there are some room for improvements that can be made to enhance the performance of the robots. For PR and TR, an external sensor can be used to allow the robot to have feedback on its current location. Based on that information, the localisation can be carried out more accurately. The TR also can be built with lighter materials or a more environmentally friendly material to reduce waste. The kicking mechanism is also bulky and complicated in designed which required a very high accuracy in dimension and positioning. 227

TEAM MIU TEAM MIU FROM MANIPAL INTERNATIONAL UNIVERSITY Hidesh, Logeswaran, Shahid, Tharindu, Rashmika. Electrical and Electronics Department, School of Engineering, Manipal International University, Nilai, Negeri Sembilan, Malaysia. ABSTRACT The theme of this year’s ROBOCON competition was the game of rugby 7’s, where the robots have to perform the tasks of passing, placing and kicking the rugby ball. Points are awarded for a successful pass, place and kick and the winner is decided upon who gets the maximum amount of points in 3 minutes. The robots were designed with the intent to be efficient and fast. The robot’s functionality, weight and safety were taken into top priority. 3D-modelling software was not used to simulate the robot’s functionality as both the robots were finished within a short span of three months with limited members. Crude versions of the mechanisms were built and tested. The kicking process of the robot was the most enduring one, as the mechanism was tested several times using different methods. Springs or surgical bands (elastic) were chosen to provide power for the kick. The use of a single motor or a powerful pneumatic piston was not used as it would require a lot of current and a lot of air, respectively. Several rudimentary designs were made and tested for the passing and placing mechanisms using wood and scrap metal. Finally, pneumatic cylinders were decided to be used as they provide excellent power transmission. A simple gripper was fitted for gripping the ball. A small pneumatic cylinder was fitted under the throwing arm for picking up the ball. For the placing mechanism, a wide variety of mechanisms like grippers and trays were tested. In the end, a slider with a slip tray was chosen as the final mechanism. For the robots’ drivetrain, four DC motors were used with the combination of mecanum wheels. Our robot is unique as it is fast and has a higher success rate. The materials used are also environmentally friendly since most of them are scrap parts. The Try Robot is unique as the placing mechanism can be used for sorting items on a production line. This sorting process would be suitable for small objects. 228

1.0 INTRODUCTION TEAM MIU The first step to make the robots is to build the chassis. Steel box bars were used to make both the robot’s chassis as they are lightweight and relatively strong. Aluminium bars or Aluminium profile bars are better options but those were eliminated due to accessibility reasons and equipment restrains in our labs. The usage of these bars would have also been quite expensive. In order to finalize the chassis design, the drivetrain of the robots must be selected. The Pass Robot and the Try Robot require omnidirectional movement in order to maximise efficiency and to save time while correcting manual driver errors. This could be overcome by the use of mecanum wheels or omni-wheels. The wheels must be strong enough to handle heavy weights as the robot’s weight may cross 15 kg. Four DC spur geared motors are chosen for these robots as they have higher speed and torque, and also require less current relatively. The use of brushless motors could have produced better results but it has a major disadvantage of requiring power supplies with higher discharge ratings which are quite expensive. The next challenge to face is the task of passing the ball to the Try Robot. The first step to pass the ball is to pick-up the ball from the ball rack. Air cylinders were chosen because of their compact and reduced control systems. Whereas stepper motors would require special driver boards and would draw more current. The usage of servo and stepper motors in both robots were eliminated due to this reason. The task of passing the ball was also solved only by using a more powerful cylinder with a bigger bore and supplied with higher input pressure. For receiving the ball, a simple basket like frame was built with nets to catch the ball. Certain parts of the Try Robot’s frame were built out of wood aiming to make the robot move faster by reducing its weight. Numerous methods of placing mechanisms were examined and the most efficient one turned out to be the sliding mechanism. The Final main task of the Pass Robot is to kick the ball. We chose the zone which delivers highest points for a successful kick and it was approximately 8 meters away. Pneumatics with the combination of springs were considered for this task. Upon testing, it was calculated that it would require a large volume of air to carry out kicking three balls since the other pneumatics system for passing was on the same robot. Hence, we searched for high torque motors to replace the piston. The wiper motors present in old cars were decided upon since they have better torque. For the control systems of the robots the Arduino board and other related components were selected to give out instruction to the mechanical parts. Custom Building Circuit Boards would have been a much more effective and cheap method to build control systems but our labs in the university lacks these equipment. 229

TEAM MIU 2.0 DETAILED DESIGN 2.1 MECHANICAL DESIGN OF ROBOT The mechanical design for the two robots is separated into four sections described in section 2.1.1 to 2.1.4. 2.1.1 CHASSIS One-inch square box bars with the thickness 1.2 mm were chosen for making the chassis. These bars are relatively strong and lightweight and fits perfectly for our current requirements. These box bars were welded together rather than using fasteners in order to maximise strength (and torque transmission in some cases). They were welded on a Metal Inert Gas (MIG) welding rig with lowest voltage settings as they melt due to thickness of the bars. 2.1.2 LOCOMOTION For the robot’s drivetrain, four DC geared motors were used with the combination of mecanum wheels for each robot. The motors used were fitted with a square gearbox with spur gears. This allowed the motors to have a high RPM which can withstand higher torque at the same time. These motors were fitted on the robot with the help of L-clamps which where custom made in the lab and welded onto the chassis. Since there is no possibility of backlash to the motors, the wheels are directly attached to the motors with custom aluminium couplers made using the lathe in the lab. Mecanum wheels were used for both robots as they allowed the robots to move in any direction. The side to side movement was crucial as it would save time in picking up the ball in the passing task. 2.1.3 PASSING MECHANISM Figure 1: Gripper claw 230

TEAM MIU Figure 2: Process diagram of gripper arm movement Powered by three pneumatic air pistons, the robot’s aim is to efficiently grip, lift and throw the ball to the Try Robot, which is approximately 1.7 metres away. Piston A as shown in Figure 1 moves horizontally in order to grip the ball from the ball rack. Followed by Piston B where it lifts the gripper’s arm vertically to level it before launching the ball. Piston C completes the process by pushing the gripper’s arm where the arm is lifted and rotated to a 70° angle to allow a smooth and accurate pass to the Try Robot. By adding the speed controller, we are also minimizing the amount of impact caused to the base of the gripper, which allows the robot to have no damages or scratches during the game. The average time taken to throw the ball by completing the three tasks above is an outstanding 2 seconds. This makes our robot lethal in terms of speed and accuracy compared to the other competitors. The process diagram of the gripper arm’s movement is shown in Figure 2 for better understanding. 2.1.4 CATCHING & PLACING MECHANISM Figure 3: Slip tray The unique thing about our Try Robot is that it is made out of lightweight wood and steel material which makes the robot extremely light (10 kg). This highly enhances the movement speed of the robot, covering 1 meter in under 5 seconds. A box with slide-like 231

TEAM MIU nets attached to all four sides acts as a catching mechanism for this robot. Once the ball is passed to the Try Robot, the nets provide a soft landing allowing the ball to roll into the “Sliding Tray” as shown in Figure 3. Piston B activates after this process by pushing the “Sliding Tray Frame” 30 cm outside the robot and into the placing rack zone as shown in Figure 4. Once the “Sliding Tray” is in position, Piston A is activated, where the lower net attached with bar is pulled out, allowing the ball to fall into place. The sliding tray frame is then retracted to its initial position slowly. The whole process takes an average time of 3 seconds. The process diagram is shown in Figure 4 for better understanding. Figure 4: Process diagram of Try Robot’s sliding mechanism 2.1.5 KICKING MECHANISM Figure 5: Kicking Mechanism 232

TEAM MIU Figure 6: Process Diagram for Kicking Mechanism The kicking mechanism design is inspired from a catapult design. Two single way bicycle gears connected to motorbike sprockets are attached to each side of the axle. These sprockets are chained and are connected to two car’s wiper motors which pulls the chain, where the rotation of the axle takes place automatically as shown in Figure 5. Two hammer springs are attached from the ends of the axle to the base of the robot creating force when the axil is rotated. The kicker is capable of kicking the ball up to an average distance of 10 meters length and 2 meters height. This puts our robot at an advantage as we will be kicking the ball from Kicking Zone 3, scoring 20 points for each try. Figure 6 shows the working process of the kicking mechanism in detail. Why is this ABU ROBOCON 2020’S winning robot? • Outstanding 2 seconds to grip, lift and pass the ball • Quick and accurate placement of ball into rack in 3 seconds per try • Able to score points from Kicking Zone 3 • Lightweight Try Robot, increased movement speed (under 7 kg ) and covers 1 meter in under 5 seconds. 233

TEAM MIU 2.2 ELECTRONICS DESIGN The control system of each robot consists of: • Arduino Uno • The Uno is a widely known developer board known across the world for its implementation in robotics. It is a perfect fit for its use in this project as it has many PWM output pins and wide variety of other functions. • MD110A Motor Driver • This Motor driver board chosen as it meets the power rating requirements of the motor and has many safety functions (eg: Reverse Polarity). Two of these drivers were used in each robot to control each motor separately • Wiper Motor Driver + Relay • Since there were two wiper motors used, a driver with high peak current rating with a combination of another relay to control each motor separately • Relay-Breakout Board • This Breakout board of relays is used to control the solenoid valves. • HC-05 Bluetooth Module • The robot is manually controlled and Bluetooth is chosen as the medium of communication. This board is chosen for its compact, cheap and effective design. Others: • 12 V geared motor • Solenoid valves • Wiper motor Power supply: Four Lipo-batteries (11.1 V, 2.2 Ah) were used for the Pass Robot and two of the same batteries were used in the Try Robot. 234

Communication: TEAM MIU Joystick components: • Arduino Nano • This Arduino board was used for its compact design. • Analog joystick • To control the movements of the robot • Toggle switches • To control 2.3 SOFTWARE DESIGN Code for joysticks: As mentioned in electronics parts, there were two objectives to meet. 1. Control the mecanum wheels 2. Control the pneumatic pistons The commands taken from the Groove joysticks and toggle switches were converted into byte and encoded string format before sending them to the receiving channel (Robots Bluetooth Module). Joystick values will be assigned to each motor driver channel as the direction and as well as the speed. These directions vary from wheel to wheel according to the moving formation required. Code for Robots: Initially, in the Try Robot, motor drivers, relays and wiper have been defined as OUTPUT pin modes. As soon as the receiver robot stop receiving the encoded string from the joystick, program in the robot’s loop function will stop all its tasks except for maintaining relays current positions. This can be identified as a robust feature to avoid robot getting malfunctions in case if the connection is lost with the joystick due to any unexpected reasons. 235

TEAM MIU 3.0 ROBOT TESTING Upon testing, many minor faults were found and rectified. Mainly while placing the ball, the front part of the chassis of the Try Robot kept hitting the border zones which at times, thus damaging the robots. Foam cushions were kept at the corner but still, the impact was high. In the end, two springs were attached in the front to act as a shock absorber which reduced the impact significantly. While driving the Try Robot from the try spot to the Receiving Zone, due to driver error, the robot scratched the walls of the boundary. This sometimes put the robot off balance and stopped its movement. To solve this problem, ball bearings were attached at the corners of the robot to ensure smooth driving along the edges of the boundary. There was also a problem with the passing mechanism where the fastener bolt used to tighten the joint weakened and bent on every pass attempt. This was solved by welding the nuts together in the joint which transmitted the power not only to the bolt but to the entire arm. This gave a consistent throw every single time. 4.0 DISCUSSION In the end, our design of the mechanisms proved to be highly effective and efficient. Our strategy to win is to make sure that we place the first four try balls as soon as we can to start the kicking task. The idea is to kick 4 balls as soon as possible so that automatically the opponent will be short of 20 points (assuming the opponent kicks from zone 3). This strategy is by far the best as it ensures victory if you place four balls faster than the opponent. Hence our main goal was to make the Try Robot as fast as possible. Our gripper was placed perpendicular to the kicker which helped to pick-up the ball faster. This led the Try Robot to travel less distance which meant less time required. The main strength of our design is simplicity and less moving parts. This reduces the chances of mechanical failures during the competition significantly. 5.0 SUSTAINABLE ENGINEERING PRACTICES Our robots were developed sustainably. We have used recycled steel for our robot. The gearbox, chain and wiper motor used for the kicking process are taken from an old bike in a spare part shop. For air storage, we used the recycling plastic bottles. There are a couple of parts where we used recycled wood. By using recycle items we have tried to minimize the use of materials. This way, it will help to protect our environment. At the same time by using recycling items, we are also able to cut the cost of our project. We designed the robot precisely to not waste material. The same materials can be used for the next competition if needed. 236

Finally, we can say that our project is sustainable for an environment, not harmful TEAM MIU for human life, save the resources and the technology we have used in this project can implement or improve for the quality of life for everyone. We believe that even the smallest conservations we make today can bring a huge change in the future. 6.0 CONCLUSIONS In the end, the entire team was happy with the way we made our robot with our given resources. If given more time and budget, we would have built a record setting robot which could have competed in the international level There are both major and minor limitations of the robot. Since the team had a low budget, we could not buy motors which could handle heavy weights and could travel at higher speeds. This eliminated our opportunity to automate the kicking process (as this mechanism is heavy) which would significantly save time. The kicking process could be automated by building mechanisms in the Pass Robot which allows to hold three to four balls then automatically place them on the ground with the tee to kick. The other major drawback is not automating the robot. Again, due to our budget, time and having just 3 active team members we could not implement advanced simulation and automation systems. For future competitions similar to this, we would recommend to do a 3D model complete version of both the robots and plan the budget accordingly to increase speed and efficiency of the robots. If possible, we would also recommend buying higher grade dc motors with built-in encoders and planetary gearboxes for better performance. Finally, if the resources are available, we would recommend to automate the robot with sensors and cameras in combinations with advanced ROS systems to find the shortest distance required to complete the tasks and also drive efficiently to ultimately finishing the tasks in the shortest time. 237

REC REC FROM UNIVERSITY OF MALAYA Wong Hong Gao, Wong Yong Liang Department of Electrical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia. ABSTRACT The two robots being built are inspired by the ABU ROBOCON 2020 theme which is “ROBO RUGBY 7s”. The competition is based on the rules of the rugby game, which involves gameplay elements such as kick, pass and try. Two robots will be required in this challenge. Pass Robot (PR) which is being used to pick up Try Ball from the ball rack and pass to the Try Robot (TR). After receiving the Try Ball, TR will transport the Try Ball to across five obstacles and reach the Try Spots to score a Try. For every successful try, we will be able to obtain a Kick Ball. PR will be used to kick the Kick Ball which is being loaded on a tee. The main challenges to be solved are the locomotion, try, receive, pass and kick mechanisms on the robot. A few more aspects that we look into would be the structural design and also error control mechanism which will be discussed later on. In terms of reference designs, we used inspirations from previous ABU ROBOCON contests. In order to validate designs, computer simulations and also prototyping models were being created and being improved on. Throughout the design period, most of the time is being spent on testing and optimizing the performance of the robot. 238

1.0 INTRODUCTION REC There are two robots being designed which are the Pass Robot (PR) and Try Robot (TR). Both robots are made from aluminium profiles which eases the building process. mecanum wheels are being used to allow omni directional movement [1]. Both the robots are being controlled manually for locomotion and other competition objectives. TR is designed to complete the pickup of the Try Ball and transporting of the Try Ball to Try Spots. PR is mainly focused on passing the Try Ball to TR and kicking. Pneumatic system is being implemented on the PR for the pass process. A gear-motor system is being utilized to initiate the kick process. The MCU used onboard is Arduino Mega which is used to control the locomotion system, kick, pass and try mechanisms. Both robots are each powered by two Li-Ion battery packs with a nominal voltage of 22.2 V and a distribution board is used to control the power supply. Simulations and calculations were being done before carrying out the building processes. Potential real-world applications would be in retrieving items from hazardous regions. 2.0 DETAILED DESIGN 2.1.1 MECHANICAL DESIGN ON TRY ROBOT (TR) (a) (b) (c) Figure 1: (a) View of TR (b) mecanum wheel (c) Repositioning & try Mechanism 239

REC The TR resembles the shape of a cuboid with a ‘funnel’ like top structure. The main structure of TR is made from aluminium profile, which is easier to make connections and joints. mecanum wheels are used for enabling omnidirectional navigation, which makes it easier to navigate the ‘obstacles’ on the playing field. The top portion of the TR is being covered by the PVC net and PVC corrugated board to reduce the impact force when the rugby ball hits the TR. The funnel-like design provides a larger area for the rugby to be captured and funnelled down to the Try mechanism. The Try mechanism can be found at the bottom. It consists of a draw-bridge style system that allows the rugby to be lowered. Also, a roller is present to reposition the rugby to initiate a proper Try. After the rugby is properly positioned, the roller will squeeze it through the opening at the bottom part of the TR in order to complete a Try. Figure 1 shows the various parts of TR. 2.1.2 MECHANICAL DESIGN ON PASS ROBOT (PR) Figure 2: (a) View of TR (b) Compressed air tanks in pneumatic system (c) Gear-motor system The PR handles both the Try Ball pickup and passing mechanism, and kicking mechanism. PR is shorter when compared with TR, to make the robot more stable. Similar to TR, its main structure consists of aluminium profiles. The locomotion system is similar to TR. A pneumatics system is used for the Try Ball passing mechanism. Then the weight is released like a pendulum, with the help of gravity it will fall and hit the kick ball, hence, kicking the A fork-like design is used to pick up the Try Ball and catapulted by the action of the pneumatic cylinder. The gear-motor system is driven by a free-gear 240

unidirectional motor, which raises a weight to a certain height of the ball. The kicking REC mechanism is also coupled with springs to increase the stored potential energy in the system, providing more kicking force. Figure 2 below shows the various parts of PR. 2.2 ELECTRONIC DESIGN Figure 3 shows the locomotion circuit schematics for both robots: Figure 3: Schematic for locomotion components. *Note: the ibt-2 LPWM and RPWM pins are connected to the Arduino digital pins for control. IBT-2 H-Bridge Motor drivers used to power the movement motors of both robots with a maximum rated current of 43A. The movements of the robots have to be fast. Thus, having high amperage is important for this application [2]. L298N A dual H-bridge motor driver that powers the lifting motor and the repositioning motor. With a maximum rated power of 25 Watt. Since the motors controlling the lift and the repositioning mechanism does not have to move fast. The peak power delivery is sufficient for the purpose of the motor driver [3]. 241

REC Figure 5 shows the TR specific schematics: Figure 5: TR specific components LM2596 Since the power supply two 11.1 V LiPo batteries connected in series. The total voltage of 22.2 V has to be stepped down using a LM2596 to be usable by the motor driver. The LM2596 take a voltage from 4.5 V to 40 V and creates an output of minimum 3.16 V and continuous 3 A [4]. Figure 6 shows the PR specific schematics: Figure 6: PR specific components Pneumatic valve 12 V pneumatic valve attached to a relay and a separate battery and controls the pneumatic piston attached to the catapult. TB6560 stepper motor driver This controls the stepper motor used to lower the hitch at the base of the catapult to secure the rugby ball into the holder. A stepper motor is used here for better precision. 242

2.3 SOFTWARE DESIGN REC Figure 7: General Movement Flow Diagram Initially the Arduino Uno was used during the development period. However, there were concerns that the lower amount of flash memory was going to hold us back as our codes are getting longer resulting in the switch to the bigger and more expensive Arduino mega. The processing power of both the Uno and the Mega are the same as they have the same CPU with the same 16 MHz clock speed. Furthermore, the larger SRAM of the Mega also gave us the freedom to implement more variables during runtime. Since the Playstation 2 controller is used to control both the robots. the PS2X_lib library is used to interface with the controller where the controller’s button status (pressed or not) are updated with the function ps2x.read_gamepad() which runs every loop cycle. General Movement (TR and PR) Figure 7 shows the general movement flow chart. The L1 button is used to toggle between two different sets of controls. The Dpad is used to create fast but inaccurate movements where the input to the motor is a constant value. The left and right triggers are used for rotational movements where a single press equals a 90° turn. The joystick is used for finer movements where the input to the motor is small and varies with the movement of the joystick. 243

TR specific functions: The right joystick controls the up and down of the lift mechanism. The score button spins the motor attached to the lift to unload the rugby ball. The reposition spins the rugby ball to make sure that the rugby ball is align with the score placement box. Figures 8 and 9 show the controller layouts on TR and PR. The final control layout is as shown below: REC Figure 8: TR controller layout Final control layout: Figure 9: Controller layout for PR 3.0 SIMULATION DATA For testing, we have conducted simulations and tests on the pneumatics system on the PR. Pressure is defined as the amount of force applied, normal to the surface of the object per unit area. Eq. (1) is used to calculate the most optimum achievable force by the pneumatic cylinder relative to the pressure applied from the compressed air tanks. The compressed air tank is able to store up to 6 bar of compressed air which is within specified limits. (1) 244