310 likes | 1k Views
Autonomous Vehicles. Derek Smith EKU Dept. of Technology Computer Electronic Networking. OUTLINE. Motivation: DARPA Grand Challenge and NAIT Introduction: The purpose of autonomous vehicles and future roles Problem Statement: NAIT’s settings for the new competition
E N D
Autonomous Vehicles Derek Smith EKU Dept. of Technology Computer Electronic Networking
OUTLINE • Motivation: DARPA Grand Challenge and NAIT • Introduction: The purpose of autonomous vehicles and future roles • Problem Statement: NAIT’s settings for the new competition • Proposed Solutions: Possible answers • Results: Video and Explanations • Conclusion: New understanding and Learning • Future Work: Possible extensions to my work • Questions • Thanks
MOTIVATION • The DARPA Grand Challenge was a great inspiration to me. • Along with a 2 million dollar prize. • NAIT played a big role in my motivation as well. http://media.pbs.org/asxgen/general/windows/wgbh/nova/darpa_outt_03_220.wmv
INTRODUCTION • The issue with producing an autonomous vehicle is that everything that can, always will go wrong. • Argo was a major part of my research. • The DARPA Grand Challenge • How autonomous vehicles work properly • The direction in which they are headed in the near future.
William “Red” Whittaker Carnegie Mellon University's William "Red" Whittaker and the Red Team's Sandstorm vehicle
PROBLEM STATEMENT • The National Association of Industrial Technology • Added an autonomous component to the annual robotics competition • The robot must be able to follow a 1 ½ inch strip of reflective aluminum tape • Creating a need for understanding autonomous vehicles
PROPOSED SOLUTION • To Solve this I went through many possibilities: • Calibrating and adjusting the infrared detectors I have. • Purchasing close proximately infrared detectors. • Creating a light sensitive circuit that reflects light off the tape • Creating a circuit sending voltage through the tape and receiving the information
Proposed Solution • My solution was to create the circuit that ran voltage through the tape. • Other methods need constant revision • The circuit method is simplistic • It only deals with one main variable • Less things to go wrong
RESULTS • It worked, not quit to where I wanted it, but it still works. • Problems that resulted: • At higher speeds handling has problems. • The directional decisions are entirely too slow • The bot does all the tasks that are required at this point.
Results ' {$STAMP BS2} ' {$PBASIC 2.5} DEBUG "Program Running!" voltdetector1 VAR Bit voltdetector2 VAR Bit voltdetector3 VAR Bit voltdetector4 VAR Bit DO DEBUG CLS voltdetector1 = IN0 voltdetector2 = IN1 voltdetector3 = IN2 voltdetector4 = IN3 IF (voltdetector2 = 0) AND (voltdetector3 = 0) THEN DEBUG "Moving Forward" PULSOUT 13, 850 PULSOUT 12, 650 ELSEIF (voltdetector2 = 1) THEN DEBUG "Front Left is off track" PULSOUT 13, 850 PULSOUT 12, 850 ELSEIF (voltdetector3 = 1) THEN DEBUG "Front Right is off track" PULSOUT 13, 650 PULSOUT 12, 650 ENDIF LOOP • To the right is my code • Below is the main circuit
CONCLUSIONS • Not yet ready for the DARPA Grand Challenge • We have made head way for NAIT next year • That is, if my predecessors don’t squander their time and resources. • Understanding an autonomous vehicle was the goal of this project • I have obtained that goal • I’m still in awe of the capabilities out there, but I at least know the basics of how they work.
FUTURE WORK • Possible extensions • Add vision analyses, nothing big • Create infrared mapping • And possibly add GPS • Any suggestions will be very welcome, email me at bruins_78@hotmail.com
Questions Does anyone have any Questions?
THANK YOU Thank you for your time and attention!