Helping Hand Senior Design Final Documentation
Group 10 Spring 2013
Kurt Graf
Taylor Jones
Matt Carlson
Eric Donley
4/29/2013
1
Table of Contents
Chapter 1 Executive Summary 5
Chapter 2 Project Description 8
2.1 Motivation and Goals
2.2 Project Requirements and Specifications
Chapter 3 Research Related to Project Definition 11
3.0 Introduction
3.1 Similar Projects
3.2 General Research on Robotics
3.3 Robot Business
Chapter 4 Research Related to Components 38
4.1 Relevant Technologies
4.2 Possible Components
4.3 Example Architectures and Diagrams
Chapter 5 Design Details 59
5.1 Design Architecture and Related Diagrams
5.2 Motor Subsystems
5.3 Sensor Subsystems
5.4 Wireless Subsystems
2
5.5 Controller Subsystems
Chapter 6 Design Summary 93
Chapter 7 Build Plan 97
7.1 Parts Acquisition for Prototype
7.2 Prototyping
7.3 Hardware Assembly
7.4 Software Integration
Chapter 8 Project Testing 104
8.0 Testing checklist
8.1 Testing Environment
8.2 Motor Testing
8.3 Sensor Testing
8.4 Wireless Testing
8.5 Controller Testing
Chapter 9 Administrative Content 124
9.1 Milestones
9.2 Budget and Financing
9.3 Looking Toward EEL4915 and Back
Chapter 10 EEL4915 Design Implementation 126
10.1 System Final Design Summary
3
10.2 System Construction
10.3 System Performance
10.4 System Budget
10.5 Project Review Successes and Errors
10.6 Project Conclusion do anything different?
Appendices 134
Appendix A Copyright Permissions
Appendix B Datasheets
Appendix C Software
4
Chapter 1 Executive Summary
The purpose of the senior design project, from the course description of objectives is "To provide students a complete design experience, including the necessity to set design goals and objectives, integrate knowledge, exercise engineering judgment, plan to meet a budget and schedule, to work as a team member, and to communicate in writing." To accomplish these objectives our group, Helping Hand, designed a tele-operated master-slave robot arm which motion tracks a human operator's arm motion including open-close hand tracking by the end-effector. The project is exclusively focused on the electronics and software to control an electrically operated robotic arm. Stock mechanical robot arms and servo motors were utilized as needed to accomplish the project.
1. Why study the human operated robot arm? The future of robotics in manufacturing and assembly is increased flexibility both in mechanical performance and ubiquitous integration with human workers. The future of robotics is greater dexterity, easier and quicker program-ability, and safe operation with human co-workers. Building a tele-operated master-slave robot arm driven by sensors worn on a human arm is investigating future possibilities and general performance considerations of advanced robotics. See Section 3.0 Introduction - Unimate to Baxter. 2. Initial feasibility investigations of the project included web searches for similar projects and several were found on YouTube and are illustrated in the section 3.1 Similar Projects. Other projects reviewed more closely included previous EEL4914 projects from 2006, DUCSS a DARPA autonomous vehicle sensor system, and a 2009 project Luke's Hand, an attempt at a human prosthetic robotic hand. From searching robot websites and robot shop sales/forums sites, an initial concept of the design was determined: MEMS motion sensors and a force sensitive resistor would provide the sensing of the human operator's arm and hand motions, an Arduino-type microcontroller would process the sensor data and generate signals through a motor (dc) driver subsystem (not needed with subsequent servo motors) to drive the robot arm motors. 3. Because Robotics is virtually a graduate school field of study, section 3.2 General Research on Robotics (as relates to our project) was added to connect our project to the more general knowledge base of robotics. 4. The core design of the project is the software interfacing the sensors to the motors through the control board explained in section 5.1 Design Architecture and Related Diagrams. Fulfilling the project requirement for at least one custom designed printed circuit board, the control board design accomplished the control requirements of the project and:
Contain enough I/O pins to monitor all the sensors and send data to the motors.
5
Have enough memory to store the control software and data gathered from the sensors.
Be fast enough to only have a slight delay between the operators motions and the robots motions.
5. Specifications for the project prototype include the robot arm's dynamic range (see 5.1.1 Robot Arm Kinematics showing mechanical views of the OWI arm)
SPECIFICATIONS
Assembled Size: 9in x 6.3in x 15in
Unit Weight: 658g
Battery: 'D' size x4
Definitive response times for the robot arm's motion were not determined because of severe time constraints created by starting the project software coding almost two months late (March 23). Nominally, motion-tracking of the human-operator's arm should follow within tenths of a second or better as has been observed in a YouTube video Haptic robotic arm 1 see Fig. 3.1.3 in section SIMILAR PROJECTS. Our system does not respond this fast; however, we are driving six rotating links and not two as in Fig. 3.1.3. The initial plastic mechanical arm with small dc motors was able to track steady rotation of the gyro sensor very effectively in time and proportional motion response through two degrees of freedom of the base horizontal rotation and shoulder link vertical elevation. Only one 3-axis gyro (STM L3G4200D) was acquired for the prototype, since 6-axis (InvenSense mpu-6050) motion sensors were chosen for the final metal arm (no SPI interface to connect two more sensors to the prototype microcontroller), the prototype development was not pursued any further. A note here - that human response time to external stimuli is generally considered not faster than 0.10 second.
6
6. A first milestone was achieved Nov. 4 with the successful build of the robot arm OWI 535 EDGE (see Fig. 4.2.1) which is made from all plastic parts except for the motors and self-tapping fastener screws and has a simple manual controller arrangement - double throw switches that are manually thrown (+/- 3 Vdc) to rotate clockwise and counterclockwise the five dc motors of the 535 arm. Female jumper pin wires from each dc motor were easily accessed to connect our own motor drive system.
7. The low budget estimate of $400 was raised to nearly $800 (see section 10.4 System Budget) by the completion of the project. The OWI arm only cost $40, but it only provided proof-of-concept of our electronic control system. For presentation level of performance, we upgraded (see section 4.2.0 H/W Upgrade Options) to a Lynxmotion AL5D metal arm for $130 with elbow and wrist rotation upgrades costing an additional $30 each and seven servo motors costing about $100 plus two (large base and shoulder) replacement servos at about $100. Another cost overrun was that even though the 6 axis sensors (gyroscopes/accelerometers) were acquired from Ebay for $20 for four (mpu-6050) sensors, the pcb cost was $30 and then a stencil had to be acquired for $10 to surface mount solder the sensors onto the boards, plus miscellaneous items needed for construction (board components), plus shipping costs, etc. It is a significant point of comment that even at minimum wage per hour cost allocation for design and build of the printed circuit boards, there is little if any savings over purchasing the boards directly from vendors. For instance, the three 6-axis sensors boards could have been purchased for $50 each and $10 for shipping for total cost of $160. The basic component costs of the self-made sensor boards was not less than $70 (including shipping), even at only three hours total for design and build (very low compared with our groups experience) the total cost is easily over a $100 (travel costs to campus to use the radio club smt machine, $10 contribution to the radio club, more etc.s). A very similar calculation applies to the custom construction of the mcu (ATxmega1284AU) used in the project. 8. The build plan for the project was already started, as mentioned above, with the successful build of the plastic OWI robotic arm. Final prototype electronic parts selection was derived from the control board design and component acquisition was begun in late December and early January. Details for building the prototype and final Helping Hand are in Chapter 7 Build Plan. A list of components acquired is also in Chapter 7. Although the metal arm was acquired by the end of January, a decision to delay the build while acquiring the elbow rotation assembly and to implement the final 6+1 design directly was a severe mistake. Also, using servos that did not have the robot arm manufacturers specific (bracket mounting) compatibility probably was not a good idea. Then, non-delivery from a vendor (servos), created a combined seven week delay in project progress. The last month of the project was done on an as-needed basis and testing was done purely on a go/no-go basis while building. Fortunately, component integrity held up well.
7
The sensor subsystem consists of three Invensense mpu-6050
