Concept Generation

Robot Soccer ECEn 490 - Senior Project

WALL-E Jacob White Jordan Hofhiens Greg Stewart Christopher Knell Table of Contents Table of Contents ...... 2

Introduction Overview...... 3 Purpose...... 3 Procedure……………………………………………………………………………………….....3

Project Requirements Summary of Body of Facts and key assumptions...... 4 Customer Needs……………………………………………………………………………….…..5 Product Design...... 5 Block Diagram………………………………………………………………………………….....6 Design and Specifications ...... 7

Concept Selection & Generation Critical design parts……………………………………………………………………………….7 Concept definition sheets/Alternative Concepts………………………………………………... Concept evaluation worksheet………………………………………………………………….. Chosen Concept Design………………………………………………………………………....

Summary...... Introduction

Overview

Robot soccer is a senior project originally done by BYU over ten years ago. The project, historically, brought crowds of people to see robots play soccer by themselves. The project was canceled for several years, but now the BYU electrical engineering department brought it back to gain new insight into new designs. The final product will be two sets of robots playing soccer that try to score the most amount of goals in a single game. The robots must autonomously work together in order to make decisions to compete and score points.

Purpose The purpose of this document is to present all potentially viable design concepts and provide our rationale for the selected designs. Each of the designs documented below have extensive impacts on how our team will develop the final product, and especially on the AI techniques that will be used to form plays. Language and descriptions for these designs were crafted to be easy to understand and approachable for those newly familiarizing themselves with the Robot Soccer Senior Project.

Procedure We have outlined the main concepts required for our robot below. Great care was taken to list options and the metrics they would be judged on. Additionally, a weighting system was used to ensure that more important metrics were given priority. Design Matrices were used to explore possible design options and score them on their ability to fulfil and exceed specifications. Final design decisions were chosen based off of matrix scores and explanations for these decisions are documented as thoroughly and concisely as possible. Critical Assumptions: ● Hardware costs within budget

● All hardware will be able to communicate - WiFi, ODRIOD, camera, RoboClaw

● The battery will be able to provide the needed power/current

● Simulations will approximate real life

● Processor will be able to run all of the needed software

● OpenCV will have the needed performance

● All provided hardware will function according to specs

● The motors will be able to handle the weight of the robot

● We will be able to estimate the state of the game in real time from the camera

● Controls feedback loop will run in real time

Summary of Body of Facts: ● Robots must fit in a can with an 8-inch diameter and 10-inch height. All robots must be fully autonomous using on-board resources. ● The ball is a standard golf ball, with the color to be determined by majority vote. ● The field will be five feet wide by 10 feet long. The goals will be two feet wide and specially marked with a unique color. The sides of the field will be angled so that the ball cannot get stuck against the sides or in the corner. ● Each team will be assigned one of two fixed colors for each game. All players on that team must be able to wear the assigned color, and that color must be clearly visible from all angles at player height. All players on each team will be marked in the same way, so there will be no designated goalkeeper. ● Robots must be designed and operated in such a way that they do not damage other robots, the field, or human spectators. ● The offside rule will not apply. ● It will be deemed a violation if a robot drops parts on the field. ● Robots are not allowed to fix the ball to their body, or encompass the ball in any way that prevents access by other players. 80% of the area of the ball must be outside the convex hull of the robot, when viewed from the side at a perpendicular angle. ● No robot can use adhesives such as glue or tape for purposes of controlling the ball or constructing a dribbler. It is a violation to leave residue on the ball or the field.

Customer Needs: Customer’s needs were divided into two main audiences EcEn-On-Display as well as our professors. For EcEn-On-Display we need to satisfy our hard deadline to make sure that we have our robots ready to present by April 16th. These robot must be able to move to specified coordinates quickly determined by the overhead camera and use plays and strategies that we have coded up to be competitive at the soccer tournament. To meet our professors needs we also need to make sure that we have utilized our opportunity for design experience, teamwork collaboration and unstructured learning. These needs must be addressed directly in a design specifications

Product Design: Based on the needs and priorities discussed our focus will be: 1. To have robots manipulate the ball to a desired position. (The Goal) 2. Robot will be able to locate its own, the balls, as well as all other players’ current positions on the field. 3. Robot needs to predict ball location. 4. Robot needs to move in any direction. In order to help us meet these goals we need to have smart, durable robots. We will discuss selecting the physical robot layout in some depth later on, but for now we will discuss how these needs are going to be obtained. The robot functional capabilities will necessitate our group dividing the work. One of us will work on the physical layout, another team-member will work on the robot vision, another team-member will work on the motion control, and the last team-member will work on connecting these pieces together for intelligent play. Once we have the robot constructed that is omni-directional we will

need to have the robot vision functional enough to transmit all the coordinate points through ROS to the motion control of the the robot. Our programmed algorithms will prescribe whether our robot should approach or go around the other objects on the fields and at which speeds. Once we have all the mechanics of the robot constructed and implemented then we can branch out and add on different functional elements to our robotic makeup. Whether it is a pneumatic kicker or an additional wheel to provide opportunity for mechanical advantage on the field.

Block Diagram: Table 1: Block Diagram of Project Layout

Design

Specifications: Based off the designs above we quantified our desires into the following design specifications.

Metric Metric Units Min/Max Ideal # 1 Design and implementation Time Hours < 1000 < 700 2 Cost per Vehicle $/ vehicle < $2000 < $1000 3 Coordinate Configuration time (vision) ms < 500 < 200 4 Coordinate Configuration Accuracy cm < 1 < 0.1 5 Robot Movement Accuracy cm < 1 < 0.15 6 Ball Movement Prediction Accuracy cm < 1 < 0.15 Table 2. Design Specification These design specification can be measured. Metric one is measured in time, metric 2 with dollars. The Coordinate configuration time will be measured within a for loop. The robot will read the balls position for ten seconds and see how many different position points is calculated. The accuracy of the coordinates and the robots movement accuracy will be measured at the same time by commanding the robots to go to specific points. These design specification will be instrumental in helping our team monitor our teams progress as well as hit our teams goals. There are however, still some variable that need to be addressed. These are defined in critical design parts.

Critical Design Parts: The critical design parts for the hardware will give us the mechanical advantage we in the tournament. We have played with multiple ideas for improving our robot, but the most important idea at the moment is deciding the construction material. Once the robot is constructed we will be able to design additional parts such as a kicker. After our prototype is constructed, we may consider optimizing wheel angles or other construction factors as well when we build our second robot.

Concept Definition Sheets/Alternative Concepts: There are many decisions to make, but one of the most important decisions at this point is what material we should use to build our robot, because all of the other construction task will depend on which material we choose. We considered plexiglass, aluminum, and 3-D printing.

Plexiglass: Plexiglass, also known as acrylic glass, is transparent, light-weight, strong, and shatter- resistance. It is about half the density of glass.

Aluminum: Aluminum is a strong, ductile, non-magnetic, lightweight metal. It is about a third the density and strength of steel, so it is a great metal for applications that need a compromise between lightweight materials and the strength of steel.

3-D printing: Prototyped in the 1980s, 3-D printing has recently become quite popular. After being around a while it is no longer as expensive as it used to be and is much more widely available. 3-D printing allows nearly any conceivable design to be constructed. Thin layers of melted plastic slowly deposited onto a surface, creating a 3-D object.

Concept Evaluation Worksheet: Criteria: We selected and weighted five important criteria to consider as we rate the concept alternatives listed above.

Easily extendible: The material we build with needs to be easy enough to modify as need be. As designs, strategies, and project requirements change, there is no doubt we will make modifications to our robot frame. This is one of the most important criteria, so we weighted it 20 out of 100.

Wire management: With many communicating components, one of the most important attributes of a good design is that it is easy to keep wires from touching and shorting things out. The connections need to be easy to access. Because we thought this was so important, we also rated this criteria 20/100.

Durability: Materials that are strong enough to support the strains required is a must. However, the forces that our structure will experience are not large and almost any good material will give us the strength we need. As a result we weighted this 15/100.

Cost: We have been given a set of materials and a stipend of $50. If possible, we hope to stay within this stipend as any additional materials we buy would either need special permission or come out of our own budgets. But it isn’t the most important factor in our design, so we also weighted this 15/100.

Time to Design: This category is definitely the most important of all the categories. As college students we must balance our time between multiple classes and sometimes part-time work. We need a design that we can finish building quickly so we can focus on making the robot actually work. As a result, we weighted this category 30 out of 100.

Chosen Concept Design: Easily Extendable: Plexiglass is definitely the most easily extendible material out of the three choices, because you can easily drill holes into it and create a multi-level robot, so we gave it a 5. Aluminum is difficult to machine and takes time to bend it into the shapes needed, so we gave it a 3. Though 3-D printing is highly customizable, it is also not that easy to modify. It is somewhere in between, so we gave it a 4.

Wire management: Plexiglass and 3-D printing both allow great options for wire management, so we gave them a 5. Plexiglass’ multi-level capabilities make it easy to wire things. With enough design time, 3-D printing can be customized to very effectively hold wires in any configuration managable. Aluminum is definately not the best option, not only because it is hard to work with, but it also conducts electricity. This makes it counter productive, so we gave it a 2. Durability: Aluminum the strongest, so we gave it a 4. 3-D printing and plexiglass are also strong, but since they are both plastics, we gave them a 3.

Cost: Plexiglass is definately the cheapest, so we gave it a 4. Aluminum is a bit more expensive, so we gave it a 3. 3-D printing though can be quite costly, so we gave it a 2.

Time to design: Plexiglass is definitely the easiest when it comes to design. Simply cut out 3 layers and screw them together and you are done. So we gave it a 4. Aluminum requires a bit more creativity to get the parts to come together correctly, so we gave it a 3. On the other hand, 3-D printing is the most time intensive. A complete CAD file must be generated with the exact dimensions and specifications needed for the project, so we gave it a 2.

Decision Matrix: Weighted Plexiglass Aluminum 3-D Printing Criteria Value Score Weighted Score Weighted Score Weighted Easily 20 5 100 3 60 4 80 Extendable Wire 20 5 100 2 40 5 100 management Durability 15 3 45 4 60 3 45 Cost 15 4 60 3 45 2 30 Time to Design 30 4 120 3 90 2 60 Total 100 425 295 315

Conclusion: Considering these alternatives, plexiglass is definitely the best option. Winning ratings on almost all levels it scores over 100 more points than the next-best alternative. Below is a picture of our completed plexiglass design. Summary We want to build two robots that work together to play and win a soccer game against another similar system. The game should consists of strategies the two robots can use to work together to maximize the amount of goals scored. We hope the competition will be fun and enjoyable to watch. Our intent is to meet all design requirements specified by the customer. The final design has been created based on these customer needs, time restraints, and budget. As time permits, we will increase our efforts on strategies to win the competitions in the future.

------Document Requirements: The actual document should include the following information: · Introduction and process description. This section briefly describes the process that the team used in completing the document. · Summary of the body-of-facts. This section should highlight the key facts and assumptions that the team has gathered, upon which they are basing their design concept. · Block diagram of the proposed solution. Alternative architectures considered by the team should also be included. · Selection of which parts of the design are critical to the success of the project. · For critical design areas, include: - Concept definition sheets. Here the team summarizes the key alternatives that were considered. This should include a brief description of each concept. - Concept evaluation worksheet. This section should include a write up of the selection method used to determine the concepts that best meet the product specifications. This forces the team to articulate the selection criteria, so that others can easily see the rationale behind their decision. Note: the document should include the actual Concept Scoring and Screening matrices for key product concepts. (At least two concepts should be included.) - Chosen design concept. This is the last section, and it should summarize the design that the team will take into the actual development phase. It should clearly show why this design was chosen, and what information was used in the decision. · For the remaining, non-critical design areas, document how those areas will be designed and what technologies will be used. Summary: The information included in the Concept Generation and Selection Document could be found other company documents such as Product Architecture or Design Specifications. However, there is always a series of documents which describes what is being designed and how the design choices have been made. This program documentation becomes part of a permanent project file that can be used for company reference. This information is often used in design reviews and for program checkpoints.