2. take your JACKET OFF KELLIESCOTT KENDALLJAYSON

3. Final Presentation  Members:  Jayson Nakakura: Chassis Design and Fabrication  Kellie Murakami: Circuitry Design and Fabrication  Scott Bonilla: Algorithm Development and Programming  Kendall Kogasaka: Programming Hi I’m Scott

4. Overview of Project  Micromouse:  Autonomous programmable robot designed to navigate and solve a 16x16 square maze  Mouse must be less than 25cm x 25cm, but there is no height limitation

5. Initial Goals  Find the center by solving the maze using a flood fill algorithm  Don’t run into walls or get lost in the maze  Don’t waste time with dead ends, loops, & spirals  Speed run after initially solving the maze  Compete in the mini-competition!

6. Overview of Structure of Design and Decisions Made  Block diagram  Solving and mapping algorithms  Stepping  Chassis sliders  Sensors  Original arrangment  First Modification  Second Modification  Third Modification

7. Block Diagram

8. 1 2 2 3 3 3 4 4 4 4 4 5 5 5 5 5 5 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 9 9 9 9 9 9 10 11 12 13 M 0 13 12 13 14 2 WALLS ADDED-IN WALLS UNSEEN WALLS Solving

9. Original S-Turn

10. Modified S-Turn

11. 45 Turn

12. Program Overview  Flood-fill algorithm starting from destination  No 180 degree turns  S-turns & 45 degree turns  Speed runs  Searched cells  Wall Input on 1 st visit only  Speed up

13. Maze Cell Values  Is there left / top / right / bottom walls?  Has the cell been searched?  Flood fill value  All the info is in a single 16 bit int value.  0011 0100 0000 0111  Set by bit operations (and, or)  Get by bitwise mask and shift

14. Program Flow Diagram (1)

15. Program Flow Diagram (2)

16. Parts  Chassis: Green 2092 Plexiglas  PCB: One-sided copper  Processor: Rabbit  Bipolar Stepper Motors: Jameco 237471CB  Stepper Motor Drivers: ON Semiconductor MC3479  Side Sensors: Sharp GP2D120 IR Sensor  A/D Converter: Maxim MAX118  Batteries: 10 1.2V Rechargeables  Miscellaneous: 5V regulator, chip connectors, wire connectors, On/Off switch

17. Hardware - Chassis Front

18. Hardware – Chassis Side

19. Hardware - Chassis

20. Balancing Chassis  What to use to keep chassis balanced  Sliders?  Too big  Ball Casters  The smallest ones would fit well (9mm tall)  $20 each – YOU MUST BE CRAZY  Acorn Nuts  Jackpot  Fits well  cheap

21. Sensors  How to mount sensors?  Erector set double angle brackets  Hole distances for both sensors and brackets are in metric  Cheap  GREEN

22. Sensor Placement  First Sensor Arrangement

23. Sensor Placement  Outside sensors  Problems with reading side walls  If too close then no value  Tracking problems  Horizontal center sensor  Caused problems with alignment  Mouse tends to drift towards detector side

24. Sensor Placement  Second Arrangement

27. Sensor Placement  Outside Sensors  Still having problems  Values  Tracking  Wiring  Bending wire too much  Wire breaking off of crimps  Sensor connector hits the wall first

28. Sensor Placement  Third Arrangement

31. Sensor Placement  Emitters and Detectors at different levels  No cross signal readings  Alignment  Able to read further ahead  Outside sensors do not get too close  Consistent  Wiring  Still bent  Does not hit walls anymore

32. Additional Sensor Problems  Output readings  Low Output  Poor Voltage in connection  Fluctuating Outputs  Poor Voltage out connection  Loose ground  Sensor connector  Loose contact between sensor and its connector

33. Bumpers  Sensors overhang the chassis  Sensors hit the walls and pegs first  Bumper Solution  Shaved erector set pieces  Takes impact off of outside sensors  No more worries about sensors slamming into the walls

34. Accessing Batteries  The battery packs are held tightly between the motors and sensors  Problems accessing  Many wires in the way  Taking top off  Tension on the connectors each time  Resulting in more loose connections

35. Solution  Addition of a hinge  Hinge second plexi-glass with the erector set double angle brackets  Now we have a hood (ORIGINALITY)  Looks good  No tension  No breaking connections

36. TRIPLETS

37. Resetting  Final problem before regionals  Mouse resetting itself to the starting cell  WHAT THE?  Code?  Hardware?  Bad luck?

38. Situation  We got the s turns down  Now able to do long continuous runs  Noticed resetting  We checked the code  We checked the circuit

39. Similarities  The mouse would reset on its first run on a long path  The mouse would reset on its fourth or fifth run on a short path  Noticed that as mouse kept running, voltage regulator output increased  Starting around 4.8V  Increasing up to 5.2V

40. Solution  Mouse runs longer  Voltage regulator heats up and passes more voltage  Add a heat sync to dissipate heat  JACKPOT  Runs until batteries die  Or crashes because moving too fast  Ready for regionals

41. Competition  Held at CSU Fresno  Competed against 12 other mice  Placed 3 rd by searching 113 cells  Didn't get to show off 45s  Didn't reach the center  Congratulations, Kyle!!! (2 nd place)

42. Outstanding Problems  Broken sensor at competition  Couldn't get reset code to work  No way to physically switch between different algorithms  Breaking connections of sensors when changing batteries

43. Suggestions for Further Improvement  Directly hardwire sensors to PCB  Write new reset code  Add dipswitch to the circuit  Make batteries more accesible

44. Thank you!  Any Questions???