CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Autonomous Navigation Team John Fastabend Justin Meyer Kim Head Geoffrey Hazelett Advisor Dr. Wayne Lu Industry Representative Mr. Dave Dunning Intel Inc.
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Agenda Introduction KimIntroduction Kim Background JustinBackground Justin Methods JustinMethods Justin Results GeoffResults Geoff Conclusions JohnConclusions John Demonstration JohnDemonstration John
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Introduction Autonomous navigation computerizes real world data for navigation processingAutonomous navigation computerizes real world data for navigation processing All done without remote controlAll done without remote control Lesson to be learned: Robotics is not an easy subjectLesson to be learned: Robotics is not an easy subject
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Background A robot that finds a red ballA robot that finds a red ball Takes in sensory data from environmentTakes in sensory data from environment Makes navigation decisions based on dataMakes navigation decisions based on data
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Methods Modularized and assigned partsModularized and assigned parts Top down designTop down design Bottom up implementationBottom up implementation Verify function of individual modulesVerify function of individual modules Integrate and test systemIntegrate and test system
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Results Robot System Architecture:Robot System Architecture:
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Results Navigation Algorithm: –Based off right-hand rule for maze algorithms –Finds a wall, then follows it –Goes toward ball when visible
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Demonstration Notice obstacle avoidanceNotice obstacle avoidance Camera panning, searching for ballCamera panning, searching for ball Precision turning to follow ballPrecision turning to follow ball
CS-EE 481 Spring Founder’s Day, 2006 University of Portland School of Engineering Conclusions Autonomous Navigation is a huge successAutonomous Navigation is a huge success Using sensory data, a red ball is locatedUsing sensory data, a red ball is located Possible improvementsPossible improvements –FPGA for the vision system –Handle more complex obstacles –More memory, faster microprocessor