ENPH 253 Robot Design Proposal

Transcription

1 ENPH 253 Robot Design Proposal Team One Runor Agbaire Etienne Boivin Bryan Pawlina Eleanor Wong Submitted to: Randall Kerr Andre Marziali Jon Nakane Bernhard Zender June 17, 2014

2 Abstract Over the course of several weeks an autonomous tape-following, IR-detecting, ziplining, artifact-collecting robot is to be designed and developed for the Engineering Physics 253 competition. The robot must be able to navigate an instructor-designed obstacle course by following a path made of black tape, collecting and storing magnetic artifacts in a basket, using IR detectors to follow a beacon in the final stretch of the course, and finally using an extendable arm to grab hold of a zipline and hoist itself up, all this in as short a time as possible. The robots behaviour is controlled by several electric circuits and software algorithms, and will receive multiple inputs from its surroundings in order to guide it through the course. The roughly 7kg robot will begin by receiving inputs from two tape sensors mounted in the front which will be fed through a PID controller, which will in turn vary the power supply to the right and left rear wheels in order to attempt to follow the tape path as closely as possible. As the robot moves through the course, it will detect any artifacts along the way by using an IR detector mounted on its side to monitor the status of an IR beacon mounted on the end of an aluminum arm attached to the top of the robot. Passing by an artifact will block the beacon and prompt the robot to stop, and lower the arm which will pick up the artifact by way of a magnetic pad mounted on the end of the arm. The arm will then rotate, drop the artifact into its top-mounted basket and continue on its way. When the tape gives way to a rocky path, the robot will then begin navigating using two IR detectors mounted on the front of its chassis towards a 10kHz beacon by utilizing the signal difference detected by the sensors. Once it has reached the beacon, a final artifact (the Idol) will be collected and placed in the basket. The robot will then extend a zipline arm mounted on the top of the chassis which will latch onto the zipline and allow the robot to slide all the way back to the beginning of the course.

3 Contents Glossary List of Figures List of Tables i ii iii 1 Introduction and Overview of Basic Strategy 1 2 Chassis Main Base Basket PCB and Tinah Storage Materials and Method of Fabrication Design Specifications Order of Assembly Drive and Actuator System Steering Artifact Arm Zipline Arm Transmission Calculations and Requirements Motor Allocation and Purposes Sensors and Feedback Tape Sensors Beacon Detectors Magnet Pad Touch Sensor Artifact Detection Sensor Zipline Detection Sensor Rear Wheel Drive Control Active Control of Front Wheels Artifact Arm Servo Magnetic Pad Elevation Control Artifact Arm Up/Down Motor Control Electrical Design PCB0 - Power Rail Bar PCB1 - QRD Sensors and Artifact Sensor PCB2 - Magnetic Pad Sensor and Artifact Arm Control PCB3-2 H Bridges (Wheels) PCB4-2 H Bridges (Zipline Arm) and Zipline Switch PCB5 - IR Sensors Physical Wiring Software Code and Algorithm Overall Software and Function Flow Tape Following Algorithm Artifact Collection Processes IR Following Algorithm Zipline Retrieval Processes Error Handling Risk Assessment and Contingency Planning 25 3

4 8 Task List, Major Milestones and Team Responsibilities Major Milestones Team Responsibilities Conclusion 28 A Appendix A 29

5 Glossary Band-Pass Filter: Used to discard signals that are not within the specified band of frequencies Digital/Analog: Discrete/continuously varying signals respectively Comparator: A circuit that generates an on or off signal based on whether an analog signal is above or below a threshold H Bridge: A circuit designed to allow two different direction of current to pass through a motor IR: Infrared Microswitch: A mechanical switch that completes the circuit when depressed Op-Amp: Operational amplifier, a building block in electronics PCB: Printed circuit board Photodetector/Phototransistor/OP805: A transistor that takes IR light as its base signal PID Control: A method for controlling the robot to follow tape that takes into account the distance from the tape (P, Proportional), the time spent off the tape (I, Integral), and the rate at which the robot moves with respect to the tape (D, Derivative) TINAH: The microcontroller used in the robot QRD: A reflective object sensor i

6 List of Figures 1 Main Base Chassis Basket Design PCB and Tinah Storage Box Assembled Chassis Active Steering Design Artifact Arm Design Zipline Arm Design QRD1114 Reflective Object Sensors OP805 Phototransistor Magnet Pad Switches Zipline Microswitch Detector PCB0 Power Rail PCB PCB1 QRD Circuit PCB1 Artifact Optical Sensor Schematic PCB2 Pad Magnet Logic Schematic PCB3/PCB4 H Bridge Circuit Schematic PCB5 IR Beacon Detector Schematic Flow Chart for Overall Robot Software Flow Chart for Tape Following Flow Chart for getartifact(); Flow Chart for storeartifact(); Flow Chart for retryartifact(); Flow Chart for irfollow(); Flow Chart for Zipline Processes ii

7 List of Tables 1 Chassis Parts Motor Resources TINAH Analog I/O Allocation TINAH Digital I/O Allocation TINAH PWM I/O Allocation Table of Physical Wiring Risk and Contingency Plan Task List Schedule and Task Distrubition iii

8 1 Introduction and Overview of Basic Strategy Welcome to Team One s 2014 ENPH253 robot planning stage. We are planning on making a robot that will take the standard, but complete route through the playing field. That is, our goal is to produce a robot that will repeatedly follow the black tape, collect every artifact and the idol, and take the zipline back to the starting point. Our strategy is such because taking the zipline, although a unique process that will have to be invented independently of other tasks, is logically simple and fastest, allowing us to allot more time to carefully pick up the artifacts, and navigate reliably. In our goal there exist many challenges. However, we are to solve these problems week by week, as listed in the timeline, and do so in a modular way, such that our robot can be constructed to first accomplish one task, and then another, and so on. Once this has been accomplished, we will move on to a robot that can navigate through the rock bed and collect the final idol. At this point, we must evaluate if a zipline route is still feasible, and if not, initiate building the robot that will find a way to return to the starting position alternatively. Building a robot that is reliable and durable for testing/tuning rests a lot on a solid chassis and we will initially develop a chassis that provides optimal space usage in our 12x12x16 restriction. Beginning with the PD control we worked on previously we will first realize what coding/changes to gears/weight we need to control the robot. Essential is the ability to pick up artifacts and store them. The entire team will spend a good amount of time early on developing the magnetic pad sensor to accomplish this. We have decided to utilize the fact that the artifacts are magnetic for moving them, but not for sensing them. Instead, we have followed recommendations to create a trip beam with IR light at 1-2 above ground to see when we have passed an artifact. At this point, the robot will stop and initiate the process of retrieving the artifact. The pad sensor is magnetic and covered by a thin layer supported by microswitches. At the moment the artifact becomes stuck to the pad, the sensors will be depressed and a signal sent to our robot s TINAH board that gives more instructions on storing the artifact. This task must be repeatable. Our strategy for dealing with the rock bed is to have larger wheels in the front of the robot so that there is less rolling resistance. These wheels will begin on casters, should this be a simple solution. If in fact the wheels begin jamming in the rocks or causing other problems, we have a plan in place to implement active steering on those wheels. Our detection of the 10kHz sine wave of infrared being emitted by the robot s destination will be done through phototransistors, although this signal will need both filtering and amplification. Should the length of the rock bed be too great that we cannot keep the output signal from these circuits within the range of what TINAH accepts (0-5V), we have a plan to implement gain-switching, although it has not been determined whether we will run two concurrent circuits that differ in gain, read respectively when appropriate, or have a natural electronic system that switches the signal s path based on a comparator-transistor circuit if this is feasible. Implementing the ziplining procedure will depend on a reliable coding algorithm that positions the robot within range to grab a hold of the zipline. By week five we are working on turning the robot in a way that the open face of the hook of the zipline will approach the zipline and accept it before the robot attempts to lift itself up. Development of the zipline is planned but reliant on relatively on-pace progress in the rest of the robot, and is acknowledged as a rather final development. Nevertheless, the zipline arm will rely also greatly on manufacturing quality, as several moving parts need to bear loads on their joints as great as the robot s weight. The ziplining phase is to be executed by means of a folding two-stage arm that extends via motor so that the tip (hook) of the arm remains almost exactly above the robot s center of gravity, before a servo at the end controls a hook to grab the zipline. Lastly, a high-torque winch which was extended during the arms s unfolding will pull the robot off the ground in a straight line, freeing it to roll back down to the pad. Our alternate plans include turning around on the rock pad and finding the tape again, with help from the Beacon of Hope. Our strategy is completely centered on reliability and effectiveness, and we will generally work only to accomplish the said tasks rather than finding a special solution to it. Granted; whatever solution achieves this will be novel enough. 1

9 2 Chassis The chassis of the robot will consist of a basket, circuit tray, zipline arm and an artifact arm mount on a two level main base. Majority of the robot chassis will be created in aluminum. 2.1 Main Base The main base of the robot will have two levels to create a larger surface area to mount parts on. The top level will be where the basket, circuit tray and other major components will be mounted on. The lower level will be where motors for the artifact and zipline arms will be stored. This area can also be used to store other small components or to store dead weight should it be needed. The arms will be mounted from the lower level and extend through the top level. They will also be mounted between the circuit tray and the artifact basket. The QRD sensors will be mounted on the angle at the front of the base, this keeps them away from the wheels and the front wheels steering shaft (See Figure 1). Figure 1: Main Base Chassis 2.2 Basket The Artifact Basket will be a 5x6x8 plastic structure. The base of the basket with be sloped away from the opening to prevent artifacts from sliding out during motion or during the zipline descent. The side towards the front of the robot will contain a flat edge. This edge will help unload the artifacts from the arm (See Figure 2). 2.3 PCB and Tinah Storage The circuits and TINAH board used by the robot will be stored in the circuit tray. It will be a 4x6x6 box with several slots for plates to go in, on which the circuits will be mounted. The sides of the tray will be open to allow the circuits to be well ventilated. We will also attach hinges to the box allowing for easy access to all the circuit components (See Figure 3). 2

10 Figure 2: Basket Design Figure 3: PCB and Tinah Storage Box 2.4 Materials and Method of Fabrication The chassis base, and the zipline and artifact arms will be made out of aluminium sheet metal and will be bent into 1x1 U-channels or 1.5 square tubes. The Circuit tray will be made out of aluminium sheet metal and will be assembled as shown in Figure 3, with the plates made from plastic. The sheet metal parts will be cut using the waterjet cutter, snips and a shear machine. 3

11 The plastic components will be cut using laser cutter. 2.5 Design Specifications The chassis is made from simply designed parts and a symmetrical base which makes it simple to modify should it be needed. The second level allows for motors and other components to be stored underneath the arms where there is more space. The second level will be held up with five support beams. The assembled robot will be approximately 10 inches tall and the chassis will weigh around 5 kg. This allows for six inches of extra vertical space that could be used to increase the size of the wheels or to change the starting angle of the zipline arm. The parts will be connected with bolts to allow for easy assembly and disassembly. The assembled chassis will look as in the figure below (See Figure 4). Figure 4: Assembled Chassis 2.6 Order of Assembly The chassis will be assembled starting with the chassis base,wheels as they are the main supporting structure on which the other components will go. The zipline and artifact arms and their mounts will be assembled outside of the main chassis. This allows for them to be tested independent of the rest of the robot. The arms, basket and circuit tray can be installed in any order. The following table displays the list of our chassis parts and how we are going to add them to our robot (See Table 1). 4

12 Table 1: Chassis Parts Part Name Material Method of Fastening Chassis Base Aluminium U Channels Angles Front Wheels Rubber caster wheels with aluminium frame Bolted to the chassis plate Rear Wheels Rubber wheels Drive Shaft Chassis plate Aluminium sheet metal Bolted to the chassis Base Chassis Second Level Aluminium sheet metal Bolted to chassis base plate through support beams Active Steering Drive Shaft Aluminium Bolted to caster wheels Basket Plastic Bolted to chassis plate Zipline Arm Aluminium Bolted to zipline arm mount Zipline arm mount Aluminium Bolted to chassis plate Artifact Arm Aluminium Bolted to artifact arm mount Artifact Arm Mount Aluminium Bolted to chassis plate Circuit Tray Aluminium Bolted to chassis Plate 5

13 3 Drive and Actuator System An integral portion of the robot design consists of a wide variety of components that rotate, translate, lift, grab, and propel the robot in every conceivable direction in order to complete the various tasks that have been set out for us. Most of these actions are accomplished through the use of motors and servos, linked with gears pulleys, and rigid arms that will allow us to transform electrical signals into angular displacements, and electrical power into mechanical power. 3.1 Steering The robot will have two rear wheels, each driven by a Geared Barber-Coleman DC motor. Signals from the QRD1114 sensors (in tape-following mode) or from the OP805 (in beacon-following mode) will vary the speeds of the motors driving each wheel in order to allow the robot to turn and correct itself (see PID control). Two caster wheels in the front of the robot connected to each other will also provide a secondary active steering component to the drivetrain. A servo located between the two caster wheels will vary the orientation of the front wheels with respect to the motor speeds of the rear wheels so as to provide smoother turning (See Figure 5). Figure 5: Active Steering Design 3.2 Artifact Arm The artifact-retrieval arm has been designed to move in two dimensions in space. A servo will be used for planar rotation of the arm to allow the arm to be moved into position to pick up each artifact, as well as allowing the arm to transfer artifacts into the collection basket. A GM7 DC motor will be used for up-and-down movement of the magnetic pad at the end of the arm by powering a pulley located inside the arm connected to the magnetic pad, in order to actually pick up each artifact. A magnetic pad may be lowered onto the top of the artifact where the resulting squashing effect between the artifact and the pad can trip one to four microswitches that lie between layers of this pad. Thin enough to permit the artifact to stick, but rigid enough to transfer this force to the switches, a material such as plastic may be selected. The arm then goes through its 6

14 vertical motion and angular motion until the pad holding the artifact enters the door of the basket, and continues just past the edge of the basket so that the artifact will be knocked off into it. The artifact arm will then return to its position 90 degrees from the direction of travel and this process will begin again when the next artifact trips the IR light beam (See Section 4). One rotation of a gear (spool for pulley) that is 1.6 inches in diameter will provide 5 inches of vertical travel, which is adequate in lifting the approximately 3 inch tall artifact into the storing basket. See Figure 6 for proposed design of the arm. Figure 6: Artifact Arm Design 7

15 3.3 Zipline Arm In order to conform to the size limits set in the rules of the competition, the zipline arm will start off folded against the top of the robot. Once the idol has been retrieved, an ungeared Barber-Coleman DC motor at the base of the arm will drive a system of gears in the arm, making it fully extend. A servo located at the tip of the arm will then rotate a hook to attach to the zipline, securing the robot in place. A geared Barber-Coleman DC motor will then drive a winch in the arm to lift the robot off the ground, allowing it to slide down the zipline to freedom. The zipline arms hook-like appendage on the end of its higher half is designed to reach over top of the zipline and transfer the weight of the robot to a wheel that may guide the robot down to the starting point with little friction. The hook is positioned using the servo on digital output, because the hook will only have two positions: active and inactive, and is initiated at the time that TINAH receives information from the zipline arm sensor (See Section 4) that the robot is underneath the zipline. The servo motor will turn the arm into an upright position above the zipline, and as it does this, move it through the range of a ratchet. The ratchet-like mechanism is not directly attached to the servo but to the metal of the hook itself, bearing the force that will next be described between the hook and the extending arm which will be made of metal. The aforementioned force arises from the fact that any off-center (with respect to the base of the hook/extending arm joint) portion of the hook that bears a load will be a moment arm and create a bending stress at that joint. The goal is to lock this position in place so that the wheel remains on top of the zipline. This creates a geometrical tolerance as the alternative is a more accurate placement of zipline into a wheel directly above the hook/extending arm joint. The following figure shows the design of the zipline (See Figure 7). Figure 7: Zipline Arm Design 8

16 3.4 Transmission Calculations and Requirements Assumptions: Weight distribution relatively uniform 8 curvature tape Rolling resistance neglibile Want to tranverse course in 1 min (without stopping) Given that the course involves three 8 foot long sections, a ramp, we have approximated this as four 8 foot sections. P = ωτ v = 32 ft s = 0.55 > 0.6ft s The wheels of our robot will be approximately 3 inches in diameter. 3π = c ω = v c = 16ft/s 0.25πft = 0.76Hz a n = v2 r = QRDs are placed 10 inches from pivot thus, = 1.44 ft s 2 α = = 1.2 rad s 2 τ = I α α I α = ml2 12 = 1416 (0.43ft)2 /12 = 0.81lbft 2 τ0.81lbft rad s 2 τ = 0.972lbft 2 = 1.296Nm Using a correction factor of two to be safe as an estimate, τ = 2.6Nm and the power needed will be = 1.6W. For the zipline extender, the motor torque needed is accomplished by highly gearing down the motor, which is okay because we need a small extension(approx 4-5 inches), and thus a slow one is permissible. The result is that our design already incorporates a slow moving, high torque system that raises the two halves of the arm with only the weight of themselves as the impeding force. On the other hand, the winch that will retract the entire weight of the robot some distance needs to have tremendous torque, and likely the most powerful motor. The radius of the spool can be found as follows: 7kg robot 9.81 m s 2 cm 20N = 3.43cm Hence, the max radius of the spool we will require is 3.43 cm. Numbers used are for a Barber Colman geared motor. We can do a lot better than this and at this stage in the race have no rush, as this initiates the last part of the run. We can operate the motor based on adequate torque and let the speed be something reasonable for example the winch lifts the robot 8 centimeters in 5 seconds. 2cm radius of wheel guarantees adequate torque from above, while 8/(4π 5) rotations per sec must be met. A frequency of 0.13 Hz is certainly within this motors power rating for P = τω. 9

17 3.5 Motor Allocation and Purposes The following table organizes our motor use and the accessories associated with those motors (See Table 2). Table 2: Motor Resources Motor Drives Motor Type Other Parts 1 Right Rear Wheel Barber Colman Geared Gears 2 Left Rear Wheel Barber Colman Geared Gears 3 Zipline Extension Barber Colman Ungeared Gears, Timing Belt 4 Zipline Contraction Barber Colman Geared Spool 5 Zipline Hook Servo Racheting Mechanism 6 Artifact Arm Vertical Servo Spool, Gears 7 Artifact Arm Rotation Servo Gears (may need angular) 10

18 4 Sensors and Feedback The control system of this robot is largely based on detecting and analyzing various infrared frequencies. Touch sensors will also be utilized. Feedback from these sensors will be processed through various operational amplifier circuits and translated into mechanical actuation in various parts of the robot. The rest of this section will have detailed descriptions of the other components that have TINAH board resources allocated to them. The specific connections are listed in the following tables (See Table 3, 4 and 5): Table 3: TINAH Analog I/O Allocation TINAH Analog Pin Number Device/Sensor 0 Right Tape Sensor 1 Left Tape Sensor 2 Right IR Beacon Detector 3 Left IR Beacon Detector 4 Artifact Arm Servo 5 Active Front Wheels Control Servo Table 4: TINAH Digital I/O Allocation TINAH Digital Pin Number Device/Sensor 0 Artifact Sensor 1 Pad Magnet Touch Sensor 2 Pad Elevation Control 3 Ziplne Detection Switch 4 Zipline Hook Control Servo Table 5: TINAH PWM I/O Allocation TINAH PWM Pin Number Device/Sensor 0 Right Rear Wheel Motor 1 Left Rear Wheel Motor 2 Zipline Arm Motor 3 Zipline Winch Motor 4 Artifact Arm Up/Down Motor 11

19 4.1 Tape Sensors The two front tape sensors (QRD1114 Reflective Object Sensors, see Figure 8) are to be mounted in parallel between the two front wheels, pointing directly downwards. They will be providing feedback on the robots alignment relative to the tape, and will be responsible for dictating its direction of movement. The presence of two sensors will allow the robot to more accurately follow the tape path by comparing the readings of each sensor, and adjusting accordingly. Figure 8: QRD1114 Reflective Object Sensors 4.2 Beacon Detectors The two front beacon detectors (OP805 Phototransistors, see Figure 9) are to be mounted in parallel on the front of the robot at a height equal to the height of the beacon, pointing forwards. They will be responsible for dictating the robots direction of movement once the tape path ends.the sensors will be used to detect a 10kHz signal from the beacon, and will prompt the robot to adjust its trajectory in such a way that will equalize the readings from each sensor to ensure that it is moving directly towards the beacon as it traverses the rocky terrain. Figure 9: OP805 Phototransistor A third beacon detector will exist on the side of the robot, at the base of the artifact arm. It will detect the IR signal from behind a pad sensor on the artifact arms extended part, shining back at the robot. Specifically, it should constantly detect on as long as no artifact is present underneath the pad sensor, until an artifact enters between these objects and blocks the signal. This will be the trigger for process of artifact retrieval to begin. 12

20 4.3 Magnet Pad Touch Sensor The magnetic pad sensor is comprised of 4 microswitches (See Figure 10) that are placed betweena down-facing, square, stiff plastic sheet, and a permanent magnet above them. The function of the pad sensor is to be lowered (See Section 3) to the top of the artifacts, where the magnetic attraction sticks the artifact to the pad, depressing at least one of the microswitches. 3 OR logic gates are configured such that if any of the microswitches are depressed constantly an on signal will be sent to TINAH indicating to launch the store artifact process. (See Section 5.3). Figure 10: Magnet Pad Switches 4.4 Artifact Detection Sensor To find the artifacts, we will use an IR emitter and another OP805 transistor. The IR emitter will hang from the artifact collection arm and the OP805 dedicated to artifact finding will be mounted on the side of the robot. When the signal from the IR emitter is very high, there is no artifact between the detector and the signal. Once an artifact passes through the detector and the IR emitter, signal is lost and we can retrieve the artifact that is at that position. 4.5 Zipline Detection Sensor A microswitch (See Figure 11) is to be mounted on the robot s zipline hook, below and slightly adjacent to its point of contact with the zipline. The switch will be implemented in such a way that it will be activated when the zipline hook has made contact with the zipline, prompting the winch located on the zipline arm to activate and lift the robot several inches above the ground (See Section 3). Figure 11: Zipline Microswitch Detector 4.6 Rear Wheel Drive Control The robot is to be driven by powering the back two wheels independently. Power is applied to the wheels based on the direction the robot needs to turn, which is governed by the PID control with information acquired through tape sensors (See Section 6.2). For example, if the robot is in a state too far right, more power will be applied to the right wheel than the left so the robot may realign itself. 13

21 4.7 Active Control of Front Wheels Although we are planning to test our robot s controllability on the rocks with passive front wheel casters, we have a plan to include active control of these wheels. Corresponding to the state of rear-wheel drive control, we will code a method to rotate the front wheels in unison using a servo motor. This is meant to provide the rear wheels with the least amount of friction to overcome as the front wheels should always be placed parallel to the line of travel. It has been acknowledged that in cars the two front wheels are not adjusted to identical angles during a turn, but we feel that for simplicitys sake this will be a good enough approximation; enough to allow the wheels to roll. 4.8 Artifact Arm Servo The artifact-retrieval arm s position will be controlled by a servo receiving signals from one of the TINAH board s analog pins. Its position will be dictated by a beacon detector located in the same plane as the arms point of rotation (See Section 4.2), as well as microswitches in the magnetic pad attached to the arm itself (See Section 4.3). 4.9 Magnetic Pad Elevation Control A magnetic pad located at the end of the artifact-retrieval arm will be lowered onto each artifact in order to pick them up and transfer them to the robots collection basket. A digital output signal from the TINAH board will be used to tell the motor governing this movement to move from its up position to its down position, and vice versa (See Section 3) Artifact Arm Up/Down Motor Control A PWM signal from the TINAH board will be controlling a motor governing the up/down movement of a magnetic pad at the end of the artifact-retrieval arm. When the beacon detector mounted on the side of a robot has its signal blocked, a digital signal from the TINAH will inform the motor that the magnetic pad has to be moved to its down position (See Section 4.9). Once a second digital signal is received from the magnetic pads microswitch (See Section 4.3), the motor will receive a signal to be run in reverse in order to change the elevation of the magnetic pad to its original position (See Section 3). 14

22 5 Electrical Design The proposed electrical design consists of six total PCBs mounted near the front of our robot. The PCBs will be fabricated out of circuit boards with components soldered onto them. All the PCBs will connect to a central PCB power hub titled PCB0 (See Section 5.1) through male and female headers. Many of the PCBs will also connect to the TINAH board I/O pins. 5.1 PCB0 - Power Rail Bar PCB 0 is the main PCB that all other PCBs will be connecting to. This PCB serves as a central voltage source for all the other electronics within our robot. The batteries will connect to this PCB through male and female headers. The approximate size of this PCB will be 4 inches by 2 inches (See Figure 12). Figure 12: PCB0 Power Rail PCB 5.2 PCB1 - QRD Sensors and Artifact Sensor PCB1 contains the electronics for the QRD tape following sensors as well as the circuitry for the IR emitter and detector artifact sensor. This PCB will connect to the QRD tape sensors (See Figure 13), the OP805 phototransistor detector on the side of the robot as well as an IR emitter near the end of the artifact arm (See Figure 14). This PCB will be powered with 5V, -9V, 9V and GND. The approximate size of this PCB will be 4 inches by 2 inches. Figure 13: PCB1 QRD Circuit 15

23 Figure 14: PCB1 Artifact Optical Sensor Schematic 5.3 PCB2 - Magnetic Pad Sensor and Artifact Arm Control PCB2 houses the logic gates IC chips transforming our four pad magnet switches into one signal (See Figure 15). The wiring and circuitry involved in controlling the artifact arm servo which rotates the artifact arm will be found in PCB2. This circuit board will connect to the artifact arm and will be powered with 5V, 9V, -9V and GND. The approximate size of this PCB is 2 inches by 1 inch. Figure 15: PCB2 Pad Magnet Logic Schematic 16

24 5.4 PCB3-2 H Bridges (Wheels) PCB3 will house two out of four of the H Bridges dedicated to the the rear wheel drives as well as two voltage comparators from the TINAH board. This PCB connects to the two rear wheels and to TINAH PWM outputs 0 and 1. This board will be powered with 16V and will also connect to the ground line in PCB0. The size of this PCB is 5 inches by 3.5 inches. The schematic for the H Bridge circuits can be found below (See Figure 18). Figure 16: PCB3/PCB4 H Bridge Circuit Schematic PCB3/PCB4 H Bridge Circuit Schematic from PCB4-2 H Bridges (Zipline Arm) and Zipline Switch PCB4 contains the other two of the four H Bridges and these H Bridges will control the motors in the zipline. In addition to the H Bridges, this PCB will also contain the circuitry associated with the switch on the zipline. This circuit will be powered with 16V, +9V, -9V and GND. TINAH comparators of the H Bridge (See Figure 18) will attach to the TINAH PWM I/O pins 2 and 3. This circuit will be approximately 5 inches by 3 inches. 5.6 PCB5 - IR Sensors PCB5 contains two identical circuits dedicated to the IR beacon sensors on the front of the robot that will be used to follow the Path of the Robot. These IR circuits contain a OP805 phototransistor, a DC block, an amplifier, a filter and a peak detector (See Figure 17). The 17

25 outputs of the IR sensor circuits will connect to TINAH analog inputs 2 and 3. This board will be powered with +9V, -9V, +5V, and GND. The approximate size of this PCB is 5 inches by 3 inches. Figure 17: PCB5 IR Beacon Detector Schematic 18

26 5.7 Physical Wiring All external wires associated with artifact collection will be purple. Driving and tape following wires are green. Zipline related wires will be brown and IR sensor driving wires will be gray. Power line wires will be yellow, orange, blue and red depending on what voltage they will be carrying. Other colours will be used to clearly identify processes and systems in the robot (Figure 6). Table 6: Table of Physical Wiring From To Signal Carried Colour 16V Battery PCB0 Voltage (16V) Yellow 9V Battery PCB0 Voltage (9V) Orange 9V Battery PCB0 Voltage (-9V) Blue TINAH PCB0 Voltage (5V) Red QRD Sensors PCB1 Analog Green Artifact Phototransistor PCB1 Artifact detection signal Purple Artifact IR Emitter PCB1 IR Light Purple PCB1 TINAH (Analog 0 and 1) QRD Sensor Analog x2 Green PCB1 TINAH Artifact Detection Digital Purple PCB2 TINAH (Digital 1) Magnetic Pad Switches Digital Purple Magnetic Pad Switches PCB2 Artifact On Magnet Digital Purple TINAH (Analog 4) PCB2 Artifact Arm Servo Red PCB0 PCB3 Voltage to H Bridge Green PCB0 PCB4 Zip Line Motors (Raising) Brown PCB3 Right Rear Wheel Power and Direction to Motor Green PCB3 Left Rear Wheel Power and Direction to Motor Green TINAH (Analog 5) Servo Front Wheels Active Front Wheel Control Green TINAH (PWM 0) PCB3 Right Rear Wheel Speed Green TINAH (PWM 1) PCB3 Left Rear Wheel Speed Green TINAH (PWM 2) PCB4 Zipline Extension Motor Brown TINAH (PWM 3) PCB4 Zipline Winch Brown TINAH (PWM 4) PCB2 Artifact Arm Extension Purple IR Detectors PCB5 IR Phototransistor signal Gray PCB5 TINAH (Analog 2 and 3) Analog IR detector signal Gray 19

27 6 Software Code and Algorithm Our basic strategy towards the software section is to balance modular and adaptive code with hard coded processes and values. In order to achieve the best possible performance on the competition surface and to handle any errors that arise in regards to artifact collection or zipline mounting, we will program retry steps to the artifact collection and zipline grabing processes. However, if our robot fails to be able to grab the zipline (See Risk and Contingency), we will add additional code to handle the robot returning to base through the tape following track. 6.1 Overall Software and Function Flow Figure 18: Flow Chart for Overall Robot Software The basic software flow of our robot can be split into three parts: artifact collecting, idol collecting and zipline grabbing. For all retrieval steps, we will retry at the most three times 20

28 before moving on with the the rest of the algorithm. If our robot fails to find the zipline, we will turn around and take the risky and dangerous path back. 6.2 Tape Following Algorithm In order to follow the black electrical tape track on the competition surface, we will be using a PID control algorithm to give varying speeds to the left and right rear motors. The following flow chart does not state the actually PID algorithm but describes how the robot will turn depending on what the tape sensors read (See Figure 19). Figure 19: Flow Chart for Tape Following 21

29 6.3 Artifact Collection Processes The artifact collection process will be the same for both the idol and the artifacts along the tape following track. The processes involved in picking up and storing artifacts are broken into three main functions: getartifact(); (See Figure 20), storeartifact(); (See Figure 21) and retryartifact(); (See Figure 22). Figure 20: Flow Chart for getartifact(); Figure 21: Flow Chart for storeartifact(); 6.4 IR Following Algorithm An algorithm similar to the PID control algorithm used for tape following will be used to follow the IR signal provided by the beacon. However, in the case of the IR beacon, the signal will be analog instead of digital-like. To compare the values of the left and right IR detectors, we will take the difference and translate that to a direction. The following flow chart describes in general how the robot will turn in regards to the IR signal (See Figure 23). 22

30 Figure 22: Flow Chart for retryartifact(); Figure 23: Flow Chart for irfollow(); 6.5 Zipline Retrieval Processes The zipline algorithm will involve overshooting to the right in attempts to grab the zipline and gradually move left if we fail. The following flow chart highlights how the zipline algorithm will work (See Figure 24) 23

31 Figure 24: Flow Chart for Zipline Processes 6.6 Error Handling To handle failing to pick up artifacts, we will drive backwards and forwards alternatingly three times hopefully passing through the artifact. To handle zipline error, we will overshoot to the right of the zipline and iterate closer to the left until our zipline switch is activated. In the case that are robot struggles with the Path of the Robot and the zipline, an additional loop will be added to the code that will only follow the tape to retrieve artifacts. To handle tape loss error, we will increase our proportional and differential constants such that there is minimum chance to leave the tape. 24

32 7 Risk Assessment and Contingency Planning In order to be flexible and modular in our fabrication of the robot, we will follow this risk assessment and contingency planning table (See Figure 7). Risk Likelihood of Ocurrence Unable to synchronize front wheel servo with rear wheels Rocky terrain causes IR detectors to be imprecise Artifact arm unable to pick up artifacts/idol Winch unable to lift robot off ground Robot unable to slide down zipline once attached Basket scraper unable to remove artifact from artifact arm Robot unable to navigate rocky terrain/sloped area IR/Tape detectors imprecise due to noise in circuit Table 7: Risk and Contingency Plan Impact Contingency Decision Date Low Medium Replace front wheel/servo assembly with caster wheels Medium Medium Add lenses to IR detectors in order to help direct beacon signal Low Medium Replace magnet with magnet of higher strength Low High Replace motor with one of higher power rating; Opt out of utilizing zipline Low Medium Shift weight distribution of robot towards front Low Medium Reduce strength of magnet arm Low Medium Increase torque to wheels Medium Low Add a biquad filter to circuit to reduce noise July 2, 2014 July 7, 2014 July 10, 2014 July 18, 2014 July 18, 2014 July 10, 2014 July 10, 2014 July 4,

33 8 Task List, Major Milestones and Team Responsibilities The tasks of the robot fabrication will be broken into the following sections: circuit boards, artifact collection, driving, zipline and testing. The major tasks are broken as below (See Table 8). The order of the tasks in the task list roughly correspond to the the order the tasks Table 8: Task List Section Task Time Allotted Circuit Boards H Bridges 1.5 weeks IR Sensors PCB and TINAH Storage Battery Mounts and Power Rail QRD Sensors Sensor Mounts Artifact Arm Artifact Arm Sensor System and Circuit 1.5 weeks Artifact Arm Body Artifact Arm Magnet Artifact Basket Aritfact Arm Software Driving Base Chassis 1.5 weeks Rear Wheels Front Wheels Tape Following Software IR Following Software Zipline Zipline Extension 1.5 weeks Zipline Lifting Zipline Hook Zipline Software Testing and Tuning Artifact Grabbing Testing 2 weeks Tape Following Testing IR Following Testing Circuit Testing Zipline Testing Done Done Testing 26

34 will be completed. The artifact collection section and the driving section will be developed simultaneously because these two sections depend a lot on the status of each other. 8.1 Major Milestones For July 7, the major milestones day, our goal is to be well in development for the following tasks: H Bridge Circuits IR Sensor Circuits PCB and TINAH Storage Battery Mounts and Power Rail QRD Sensor Circuits Sensor Mounts Artifact Arm Sensor System and Circuit Artifact Arm Body Artifact Arm Magnet Artifact Arm Basket Artifact Arm Software Pseudocode Base Chassis Rear Wheels Front Wheels Tape Following Software IR Following Software We will strive to finish creating majority of the PCBs and create the basic frame and mechanics for our robot. On this major milestone day, we will also decide whether or not to proceed forward with developing the zipline or to spend more time optimizing the artifact collection and tape following mechanisms. 8.2 Team Responsibilities Each team member will have their own main focus and sub focus. The table in the appendix describes how each task will be sorted into the weeks of robot fabrication (See Appendix A). Runor will be focusing on the artifact arm with sub focus as software. Etienne will be focusing on drive with sub focus as circuits. Bryan will be focusing on the zipline with sub focus artifact arm. Eleanor will be focusing on software with sub focus drive. 27

35 9 Conclusion The robot competition is to traverse an obstacle course and pick up four artifacts and return to the start of the course. There are two ways to return to the start, drive down the track used to get to the last artifact or use the zipline located at the last artifact. Overall, the robot will weigh 7 kg and will be able to pick up artifacts and use the zipline to return the start of the course. There are several things to consider when designing a robot that can complete the challenge. The first is the drive system. This robot will have two driven rear wheels with active control in the front wheels. The active control will use a servo motor to steer, this allows for reliable steering and control of the robot. The robot will use an arm to pick up artifacts. Picked up artifacts will be stored in a basket near the rear wheels. The arm will rotate between the basket and artifacts on the course. To pick the artifacts, the artifact arm will have a magnetic pad that can move up and down to move closer to the artifacts. Using the zipline to finish the course is advantageous because it is faster than driving back. The zipline arm will be at the centre of the robot, close to its centre of gravity. The arm will start folded and will extend to reach the zipline. When a switch at the end of the arm is activated through its contact with the zipline, a servo motor will move the hook into place on the zipline. The electrical design for the robot will consist of several sensors, PCBs and a TINAH board. The QRDs will be used to follow the tape on the course. The artifact and zipline arms will both have switches on them to determine if an artifact has been picked up and if the clamp is on the zipline, respectively. The PCBs and TINAH board will be stored vertically in a tray at the front of the robot. There will be two IR sensors on the robot to direct it when it reaches the end of the tape. The zipline arm and the rear wheels will have a pair of H-bridges to allow their motors to change direction. Using a modular programming approach with a schedule that will complete the minimum viable product by early to mid July, Team One s robot strives to complete the course, from following tape, to collecting artifacts, to riding the zipline out of the Temple of Doom. 28

36 A Appendix A Table 9: Schedule and Task Distrubition Week Runor Etienne Bryan Eleanor 6 (Circuit Week) H Bridge H Bridge H Bridge H Bridge IR Detector IR Detector IR Detector IR Detector 29 Final SolidWorks Design QRD Sensors 7 (Aritifact Arm and Drive) TINAH and PCB Storage TINAH and PCB Storage TINAH and PCB Storage TINAH and PCB Storage Power Rail IR Emitter IR Detector (Artifact Arm) Sensor Mounts Artifact Arm Body Base Chassis Artifact Arm Magnet Base Chassis 8 (Artifact Arm and Drive) Artifact Arm Software Base Chassis Artifact Arm Basket Rear Wheels Artifact Arm Misc Front Wheels Tape Following Software 9 (Artifact Arm and Drive) Tape Following Testing Artifact Arm Testing Artifact Arm Mounting IR Following Software Drive Cleanup 10 (Zipline) Artifact Arm Cleanup Zipline Hook IR Following Testing Zipline Software Zipline Extension Circuits Cleanup Zipline Lifting Arm, Drive Cleanup 11 (Zipline and Time Trials) Zipline Extension Zipline Hook Zipline Lifting Zipline Software Zipline Mounting Zipline Testing 12 (Buffer and Testing) Artifact Arm Cleanup Drive Cleanup Zipline Cleanup Software Cleanup 13 (Competition Week) Done Done Testing Done Done Testing Done Done Testing Done Done Testing