take your JACKET OFF

1
take your JACKET OFF

• KELLIE SCOTT
• KENDALL JAYSON

2
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 Im Scott
3
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

4
Initial Goals

• Find the center by solving the maze using a flood
  fill algorithm
• Dont run into walls or get lost in the maze
• Dont waste time with dead ends, loops, spirals
• Speed run after initially solving the maze
• Compete in the mini-competition!

5
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

6
Block Diagram
7
Solving








7
8
9
10
11
12
13
M
14
6
7
8
9
10
11
12
13
5
6
7
8
9
10
11
12
12
4
5
6
7
8
9
10
11
13
3
8
9
10
14
4
5
6
7
2
3
4
5
6
7
8
9
4
5
6
7
8
1
3
2
1
2
3
4
5
6
7
0
WALLS
ADDED-IN WALLS
UNSEEN WALLS
8
Original S-Turn
9
Modified S-Turn
10
45 Turn
11
Program Overview

• Flood-fill algorithm starting from destination
• No 180 degree turns
• S-turns 45 degree turns
• Speed runs
• Searched cells
• Wall Input on 1st visit only
• Speed up

12
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

13
Program Flow Diagram (1)
14
Program Flow Diagram (2)
15
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

16
Hardware - Chassis

• Front

17
Hardware Chassis

• Side

18
Hardware - Chassis
19
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

20
Sensors

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

21
Sensor Placement

• First Sensor Arrangement

22
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

23
Sensor Placement

• Second Arrangement

24
(No Transcript)
25
(No Transcript)
26
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

27
Sensor Placement

• Third Arrangement

28
(No Transcript)
29
(No Transcript)
30
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

31
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

32
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

33
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

34
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

35
TRIPLETS
36
Resetting

• Final problem before regionals
• Mouse resetting itself to the starting cell
• WHAT THE?
• Code?
• Hardware?
• Bad luck?

37
Situation

• We got the s turns down
• Now able to do long continuous runs
• Noticed resetting
• We checked the code
• We checked the circuit

38
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

39
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

40
Competition

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

41
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

42
Suggestions for Further Improvement

• Directly hardwire sensors to PCB
• Write new reset code
• Add dipswitch to the circuit
• Make batteries more accesible

43
Thank you!

• Any Questions???