Robotic Sensor Network Wireless Sensor Platform for Autonomous
Robotic Sensor Network: Wireless Sensor Platform for Autonomous Topology Formation Project: 04043 Sponsored By: Advisor: Dr. S. Jay Yang, CE Manager: Steven Boughton, ME Brian Teaney, CE Ryan Johnson, EE Jack Tsai, EE Matt Hrivnak, IE Gregory Rosenblatt, ME Shannon Buckland, ME
Presentation Overview • Project Overview Problem Statement, Design Process, Project Constraints • Prototype Design Overview, Testing and Problems Encountered • Final Design Mechanical and Electrical Redesign, Software Development, Final Testing • Future Plans • Questions
Problem Statement To develop a group of sensor platforms that can communicate with one another wirelessly and move from original deployment locations to form a desired network topology that offers full, energy efficient, and robust coverage. The platforms will be relatively small and lightweight. They must be able to work as a group to maximize the sensor networks life span. The end result of this project is to have a functioning group of no less then three sensor platforms that can be used as a test platform for future research and expandability.
Design Process (Phase I – Winter 2003) 1 Project Manager / Project Advisor 2 1 Needs Assessment 4 Objectives and Specifications 2 Concept Development 5 Analysis and Design 3 Feasibility Assessment Locomotion Shannon Greg Sensing Jack Matt Initial Concepts 3 Sensing Final Recommendations Jack Matt Communications Brian Ryan Mechanical Phase II Spring 2004 Final Prototype Design Electrical Hardware Movement Shannon Jack Matt 5 Sensing Communications Topology Software Brian Ryan Greg Objectives and Specifications Software Brian Ryan Greg 4
Design Process (Phase II – Spring 2004) Electronics Ryan Jack PIC 6 Brian Problems 7 PIC Brian 6 Initial Testing 8 Redesign 7 Problem Assessment 9 Final Testing Electrical Recommendations 8 Ryan Jack Mechanical Movement Brian Shannon Greg Final Design Shannon Movement Phase III Fall 2004 Topology Formation Software Topology Sensing Communications Brian Ryan Greg Mechanical Hardware Electrical Shannon Jack Matt 9
Project Constraints • Budget – Original funding fell through – Final budget not approved until March 19 • Timeline – Accelerated design time: steep learning curve – 2 phases: Prototype and Final Design • Limited Lab Equipment and Software – Limited shop time – PIC programmer – Design Software
Prototype Design Overview • 3 Tier Layout – Weight distribution – Location advantages – Ease of assembly – Ease of redesign • Locomotion Tier A – Motors, tires, power • Sensor Tier B – IR sensors • Communications Tier C – Prototype board C B A
Prototype Design Specifications • MICA 2 DOT – Run on an Atmel ATmega 128 L running at 4 MHz, w/ 128 k of program memory. – Handles the communications and network topology. • PIC 18 F 458 model – Up to 40 MHz clock. 8 k of program memory. – Will monitor the infrared sensors and control the locomotion of the robot. • Sharp GP 2 Y 0 A 02 YK IR Sensor – Less influence on the color of reflected objects, reflectivity – Current required: 33 m. A – Analog voltage corresponding to distance • Motors – Bipolar Stepper Motor – 12 VDC @. 6 A
Prototype Testing and Problems Encountered • • Turning Capability – Units could not execute turns – Solved by widening the base to 5” by 9” Programming the PIC – Initial PIC selection did not match programmer compatibility – Solved by selecting different PIC micro controller (PIC 18 F 458) Motor Control / Power – Supply currents insufficient to drive original motors – Selection of new motors and addition of higher rated regulators – Bipolar Stepper Motor (1. 8 o step, 2. 72 VDC @. 4 A, 650 g-cm) Overheating – Units were prone to overheating in a matter of seconds – Solved by adding heat dissipation elements to final design
Final Design Redesign • Mechanical Redesign – – Aspect Ratio (2: 1) Tire Coupling Sensor Mounts (4 vs. 8) Material • Electrical Redesign – – Board Size Voltage Regulation Heat Sink Battery
Final Design Software Development The final software development issue involved integration of the three main components that had been programmed separately. • Motors/Sensors • Communications • Topology Formation Algorithm
Final Design Software Integration Topology algorithm interfaces with other components • Integration with Motors/Sensors – Topology interface calls directly-related functions • Integration with Communications – Intermediate code created to link the Topology interface with proper calls to send and received messages PIC Programming Issues…
Final Testing • Hardware Testing – Movement – Errors in Turning • Software Testing – Communications Verification • PIC Testing – Unsolved Problems
Future Work • Reliable, user friendly, and upgradeable sensor platform – Programmable via wireless communication – Easy to maintain and upgrade • Scalable and robust topology formation – Distributed control leads to fast convergence – Adaptive to sensor failure or energy exhaustion • Application driven topology adaptation – Mission critical sensor network – Formation adapts to application needs
- Slides: 14