Transcript Whitewater Kayak Slalom Race Timer

Whitewater Kayak Slalom
Race Timer
Engineers:
Kevin Lockwood
Chris Munshaw
Ashley Penna
John So
Project Funded By:
Mike Neckar
Founder, Necky Kayaks
www.necky.com
Background on Whitewater
Kayaking
• Whitewater kayak slalom racing began
shortly before World War II
• This Olympic sport involves racers
paddling down a natural or man-made
rive
• Kayakers must maneuver through
hanging pairs of gates.
• Judges at shoreline determine correct
maneuvering through gates.
Background on Whitewater
Kayaking
C1 (Canoe) on a man-made course
Background on Whitewater
Kayaking
K1 (Kayak) on a natural river course
Kayak Rules
• The racer must proceed through green
gates in the down-river direction
• Red gates in the up-river direction
• 2sec penalty for touch gates but going
through
• 50sec penalty for touch and not gone
through
Present Situation
• Judge watching at each gate to make sure
the kayaker goes though
• Judge determining if each gate has been
touched
• Stop-watches used in training for timing
• Obvious problems: Human error, biases,
judges not omniscient
Our Solution
Create a automated system which tracks a
kayaker’s progress through a race course
and determines if gates are touched.
Focus on creating a reliable and low cost
product. Offset the cost of using humans
to judge gates.
Secondary goal is timing accuracy.
Marketing
• Mr. Neckar
- use for training
by olympic athletes
- introduced in races such as national
team trials (Vedder River, Chilliwack)
• Scott Shipley, US national team member
- promotion in the United States
Timeline
Overall, we are behind the proposed schedule
by about two weeks.
Our Proposed Timeline
1-Jan-07
Research
Proposal
Functional Specification
Design Specification
Assembly of M odules
Integration/Prototype Testing
Debugging/Prototype M odification
Documentation
Progress Report
16-Jan-07 31-Jan-07 15-Feb-07
2-M ar-07
17-M ar-07
1-Apr-07
16-Apr-07
Delays are caused by…
• Waiting for sensors, microcontrollers, and
RF modules to arrive.
• Testing other design options.
• Errors and bugs
• Underestimated Integration Time
• Earlier than expected deadline
Timeline
The Actual Timeline
System Overview
How to detect a Kayaker?
 Ultrasonic beam
across the gates
 RF tag triangulation
 IR beam across the
gates
Ultrasonic Beam
Advantages
 not affected by
environment
 low noise
 low power consumption
Disadvantages
 wide beam
 difficult to integrate
multiple ultrasonic
sensors due to coupled
interference
RF Tag
Advantages
 Very hard to cheat the
technology
 Low power
Disadvantages
 Difficult technology to use
 Requires a high computational
load to calculate location
 Can be expensive
Optical Beam
(Our Solution)
Advantages
 Narrow beam
 Easy to implement
 Unaffected by
environment
 Lower costs
Disadvantages
 Consumes higher power
the ultrasonic
 Sensitive to alignment
IR LED vs. Laser
• Laser (Visible Spectrum) 650nm
- coupled with a photodetector + amplifier
- very high signal strength at large
distances (5m +)
- very narrow viewing angle
- low power consumption (~20mA)
- class III and above can cause retinal
damage
IR LED vs. Laser
• IR LED 950nm
- coupled with an NPN phototransistor
- very low signal strength at distances over
2m (required amplification)
- wide viewing angle (35°) minimizing
problem of gate flexibility
- high power consumption (~100mA)
- cannot cause retinal damage
IR LED: Improving Signal
Quality
• Ambient light shielding
- used a non-reflective black paint to coat
a drinking straw (this also formed a watertight seal over the phototransistor)
• Modulation
- modulated the IR emitter with a 2kHz
square wave
- demodulating at the receiving side would
filter out noise cause by reflections of
sunlight off water, etc
IR LED: Improving Signal
Quality
• Ambient light shielding
- used a non-reflective black paint to coat
a drinking straw (this also formed a watertight seal over the phototransistor)
• Modulation
- modulated the IR emitter with a 2kHz
square wave
- demodulating at the receiving side would
filter out noise cause by reflections of
sunlight off water, etc
IR LED: Overall System
• Amplification -> Filtering -> Thresholding
- Amplification boosts the output signal
strength
- Filtering creates a steady signal
representing the amount of IR light
detected
- Thresholding creates a digital signal
representing whether or not the line of
sight is considered “broken”
IR LED: Modulation
• Decreased average current consumption
from 180mA overall to 110mA overall.
• Waveform created using an astable 555
timer
Simulation on
breadboard
IR LED: Demodulation
• Filtered using an LRC circuit, tuned to
2kHz
IR LED: Final Signal
Accelerometer
• Used to detect any contact with the gate
• 3 axis, ±5g output range
• Mounted 1 accelerometer per gate, in the
lower region of the gate (added sensitivity)
Accelerometer: Signal
Conditioning
• Low Pass Filter: allows us to “dull” the
signal and remove unwanted noise
• Comparator: gives a digital signal
representing whether or not the
acceleration of the gate is beyond an
acceptable level
-> this allows us to have the system ignore
low acceleration conditions such as gates
swaying in the wind
Accelerometer Performance Tests
• Comparator
Threshold = 1.665V
(red line in graph)
Future Improvements on Signal
Conditioning
• Have circuits printed on PCB
• Use only variable resistors reference
voltages in comparators
• Improve demodulation circuit, possibly
using an active filter
Final Sensor Signals
• Two digital signals representing the
clearance of a gate, and contact with a
gate (both fully adjustable)
• However, current consumption is
becoming high (approx. 180mA)
• This leads us to attempt ‘Presence
Detection’
Presence Detection
• Used to detect the presence of an
approaching kayaker.
• Used to trigger the turn on high power
consuming subsystem.
• Used Ultrasonic sensors
• Accuracy
• Immunity
• Ease
Presence Detection
 The sensors have an
analog output
proportional to the
distance of an object.
 Used thresholding to
detect object presence
 Used timing circuit to filter
noise.
Presense Detection
Future Upgrades
• Currently we do not have a way to
detect which direction the kayaker
came from.
• Gates are direction dependant
according to whitewater kayak
Rules.
• We will switch to IR presence
detection, due to better immunity
to environment.
• Will use one facing each direction
in gate to determine direction of
approach.
Data Communication
Requirements
• Reliable
• Long Range
• Low Power
• Fast Transmission
Data Communication Solution
 ZigBee Xbee Module from Maxstream
 30m range (upgrade 1mile)
 Current Consumption during Transmission
45mA
 UART Communication Format easy to integrate
with our Micro Controller
Data Communication
Future Updates
• We can upgrade to Xbee Pro modules for
an increased range.
• Requires more power.
• Allow software to communication back to
gates.
• Remote reconfiguration
• Remote turn on/off
MicroController Firmware
• Requirements
– Very little memory needed – Simple program
– USART Register for RF Modules
– A/D Conversion capabilities
– At least 3 inputs (IR Sensors, Ultrasonic,
Accelerometer)
MicroController Firmware
• Main Jobs
– Get a development environment running
– Integration with ultrasonic to turn on power
board
– Integration with IR sensors
– Integration with RF modules
MicroController Firmware
• Multiple Development
Environments
• 1) PICDEM
– 1st to work
MicroController Firmware
• Good Features
– Easy viewing of
ports
– Attached LEDs to
eliminate the need
to probe
– Multiple ways to
power
– MPLab
compatibility
• Problematic
Features
– Had to replace 40pin socket
– Initial running of
programs
– Quantity
MicroController Firmware
• Multiple Development
Environments
• 2) OUMEX
– 2nd to work
MicroController Firmware
• Good Features
– One LED to map
outputs of interest
to
– Programming
capabilities using
MPLab
– Less reliance on
development board
• Problematic
Features
– Building a cable
from MPLab to
ICSP
– Initial running of
programs
– Quantity – shipping
time
MicroController Firmware
• Multiple Development
Environments
• 3) Prototype
– Last and finally!!!
MicroController Firmware
• Good Features
– Cheap
– Space saving
– Easy connection to
other circuits
• Problematic
Features
– Must move to
another
development board
to program
– Determining which
components were
necessary
MicroController Firmware
• IR Flag gets
set in an
interrupt
• Accelerometer
Flag gets set in
an interrupt
MicroController Firmware
• Ultrasonic Powering Sensor Circuit
– Creates an interrupt which sets a flag
– Main program deals with this
– Output will be high when ultrasonic is high
• IR sensors Circuit
– Creates an interrupt which sets a flag
– In main program, transmission showing the
gate number and IR occurs
MicroController Firmware
• Future Improvements
– Automatic Gate Addressing
– Sleep pins on the RF module
– Polling gates for possible battery voltage
The Power
 IR sensors consume
around 150mA.
 Portable/Inexpensive
power source in a 9v
battery
 Provide clean power
at 3v and 5v for all
subsystems.
 Supply should last for
8hrs of use
Power Solution
 Isolated control directly
from Micro Controller.
 Micro Controller uses
the low power Ultra
Sonic sensors to trigger
IR sensor circuit.
 Circuit Board contains
controlled outputs at 3v
and 5v for high power,
and continuous outputs
of 3v and 5v.
Power Solution
We want our portable power supplies to
last 8 hours of continuous usage
System Power Consumption Before Power
Control
•
Total Power Required = 1.21Ahr
 System Power Consumption After Power
Control
•
Total Power Required = 0.511Ahr
Power Solution
Without a controlled power supply for 8hrs
of continuous use requires 1.21Ahr
With a controlled power supply for 8hrs
Of continuous use requires 0.511Ahr
Saves nearly 250% of our AmpHours
required.
Improves portable power supply options.
Power Solution
 We use two Rayovac 9v Alkaline
batteries in parallel for each gate
 Batteries spec at -30C to 55C
 Each Battery has approx. 0.5Ahr
Graphical User Interface
Graphical User Interface
• Purpose:
– Allows user to set up a race quickly.
– Communicates with the RF module and
collects data from gates.
– Displays data in table form.
– Automatically times the race and applies
penalties.
Graphical User Interface
• Functions:
– Kayaker list management. Add and remove
kayakers.
– Modify number of gates.
– File I/O
– Display data:
• Names
• Race Time
• Penalties applied to each gate
Graphical User Interface
• Program flow
1. User adds the names of kayakers in order.
2. User determines the number of gates.
3. User modifies the serial port settings.
• Step 1, 2 and 3 are interchangeable.
4. User presses ‘Begin’ button to begin the
race. Name list and gate number cannot be
modified from this point onwards.
Graphical User Interface
• Program flow (continued)
5. Program reads and displays data
automatically.
- Decodes gate messages sent through RF
module
- Applies 2 sec time penalty if gate touched.
- Applies 50 sec time penalty if gate missed.
6. Calculate race time and add penalties to it.
7. Table may be exported in .txt format and
uploaded to MS Excel.
Graphical User Interface
• Problems encountered:
– Exception handling
– Symbol error due to baud rate mismatch
– Repeated messages from gates
– Timing delay
Graphical User Interface
• Future Improvements:
– Time delay calculation
– Support multiple kayakers on the course
– Name list sorting
– Automatic available port detection
Summary
Created a automated system which tracks
a kayaker’s progress through a race course
and determines if gates are touched.
Focus on creating a reliable and low cost
product. Offset the cost of using humans
to judge gates.
Increased timing accuracy
The End
• Questions?
Appendix: Signal Conditioning
Appendix: Modulation
• Emitter: (Breadboard)
Appendix: Modulation
• Receiver, modulated: (Breadboard)
Appendix: Demodulation
• RLC Bandpass Filter
sCR
• H(s)= 2
s CL  sCR  1
• Using R=1, C=6.33uF, L=1mH
Appendix: Demodulation
Appendix: Demodulation
• Receiver, de-modulated: (Breadboard)
Appendix: UltraSonic Circuit
• Used a simple LM324 OpAmp with a
threshold voltage. Threshold set to
approx. 5.5ft.
• 555 Monostable Timing circuit holds
detection high for 5sec. This filters the
natural circuit noise from the ultrasonic
sensor.
Appendix: Ultrasonic Circuit
Appendix: Power Requirments
Before Power Control
 Continuous Power Consumption
• 110mA (IR circuit) + 15mA (Ultrasonic) + 25mA (Micro) = 150mA

RF Consumption
•
(150 trans. approx.@ 0.5 sec/trans) = 0.9mA
 Total Power Required = 1.21Ahr
After Power Control
 Continuous Consumption
15mA (Ultrasonic) + 25mA (Micro) = 40mA

IR Consumption
110mA (150 passes. approx.@ 5 sec/pass) =23mA
 RF Consumption
45mA (150 trans. approx.@ 0.5 sec/trans) =0.9mA
 Total Power Required = 0.511Ahr
Appendix: Power Circuit
Appendix: Power Circuit Lag
(4ms)
Appendix: Transmission
Appendix: Transmission
Appendix: Transmission Time