Capstone Design Project Capstone Design Project PlanPlan

Team SAUSAGESTeam SAUSAGESRyan Campbell

Anne Carrier

Gonzalo Gonzalez

Bryan Grider

Steve Kerkmaz

Ziad MohieddinEE 401 – EE Design I

Instructor – Dr. Mohan

April 13, 2003

IGVC StandardsIGVC Standards 0 to 8 kph speed 0 to 8 kph speed

rangerange 2 to 5 m radius turns2 to 5 m radius turns 3 m following 3 m following

distancedistance No modification to No modification to

Lead VehicleLead Vehicle Course ObstaclesCourse Obstacles

UDM StandardsUDM Standards 0 to 8 kph speed 0 to 8 kph speed

rangerange 2 to 5 m radius turns2 to 5 m radius turns 1 m following distance1 m following distance Allowed to mount Allowed to mount

targets and sensors targets and sensors on Lead Vehicleon Lead Vehicle

Obstacle FreeObstacle Free

Problem Statement and Project Problem Statement and Project ObjectivesObjectives

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Power Wheels EliminatorPower Wheels Eliminator

AdvantagesAdvantages Pre-Built SystemsPre-Built Systems Large Available AreaLarge Available Area Easily ModifiableEasily Modifiable InexpensiveInexpensive

DisadvantagesDisadvantages TractionTraction Steering ControlSteering Control

Traction ControlTraction Control Mount Conveyor Belt to Mount Conveyor Belt to

Area of Contact with Area of Contact with GroundGround

M ain point of contactw ith road

G ear

M O TO R

S teering B racket

P ivot

Chassis ModificationsChassis Modifications

Steering SystemSteering System Gear and Chain SystemGear and Chain System Servo or Window MotorServo or Window Motor

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Shuttle BareboneShuttle Barebone

InexpensiveInexpensive Interchangeable Interchangeable

partsparts Small packageSmall package 2 GB Maximum 2 GB Maximum

memory capacitymemory capacity

I/O CardI/O Card

Counter/TimersCounter/Timers Used to create PWM signalsUsed to create PWM signals

Digital LinesDigital Lines Used for UltrasoundUsed for Ultrasound

A/D convertersA/D converters Possibly used for backup sensorsPossibly used for backup sensors

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Ultrasound SensorsUltrasound Sensors

Donated by BOSCHDonated by BOSCH Once in diagnostic Once in diagnostic

mode, outputs 1 byte mode, outputs 1 byte every 25 msevery 25 ms

1 bit = 1 cm of 1 bit = 1 cm of distancedistance

Aluminum Membrane PotAluminium Membrantopf

EntkopplungsringDecoupling Ring

HülseCollet

DämpfungsschaumDamping FoamSystemträger

System Carrier

LeiterplattePC Board

KunststoffgehäusePlastic Housing

VergussmassePotting Material

SteckerConnector

CMUcamCMUcam

• SX28 Microcontroller interfaced with a OV6620 omnivision CMOS camera.

• $ 109.00

OVERVIEW

CMUcam Vision SensorCMUcam Vision Sensor

• 80 * 143 image Resolution.

• Dump images (color blobs).

•Track images at 17 frames / seconds.

• Find centroid of image.

• Adjust the camera’s image properties.

• Ability to control a servo or use one Digital I/O pin.

• RS-232 serial or TTL data communication.

MAIN FEATURES AND FUNCTIONALITIES

CMUcam Vision SensorCMUcam Vision Sensor

• Response to ambient and florescent light.

• Servo’s sensitivity to image tracking.

DISADVANTAGES

CMUcam Vision SensorCMUcam Vision Sensor

• Eyes for vehicle navigation.

• Preliminary steps for steering control.

RELEVANCE TO OUR PROJECT

IR SensorsIR Sensors

How does it work? How does it work? The black and white disk is for creating The black and white disk is for creating

pulses.pulses.

The reflected beam on the black segment The reflected beam on the black segment will cause the sensor to output a signal will cause the sensor to output a signal (pulse).(pulse).

Each pulse corresponds to a specific Each pulse corresponds to a specific distance traveled.distance traveled.

The number of pulses in a given time will The number of pulses in a given time will determine the speed.determine the speed.

SpeedSpeed Number of pulses = (n)Number of pulses = (n) Distance = (n * x)Distance = (n * x) Speed = Dist. / Time.Speed = Dist. / Time.

10 %10 %

10 %

10 %

10 %

10 %

10 %

10 %10 %

X-Axis

MovementDirection

X

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Matlab/SimulinkMatlab/Simulink

xPC TargetxPC TargetReal-Time WorkshopReal-Time WorkshopStateflow CoderStateflow Coder

xPC TargetxPC Target

HOST PC (desktop)• Development of the

algorithm• Non-real time testing

of the algorithm• Tuning of the algorithm

using real-time signalscoming from the Target

PC

HOST PC (desktop)• Development of the

algorithm• Non-real time testing

of the algorithm• Tuning of the algorithm

using real-time signalscoming from the Target

PC

TARGET PC (on vehicle)• Real-time simulation of

the algorithm• Runs the final algorithm

on vehicle• On board implementation of the

algorithm

TARGET PC (on vehicle)• Real-time simulation of

the algorithm• Runs the final algorithm

on vehicle• On board implementation of the

algorithm

Real-Time WorkshopReal-Time Workshop

Stateflow CoderStateflow Coder

Designs complex control systems based on finite Designs complex control systems based on finite state machine theory. state machine theory.

Represents the system logically and can Represents the system logically and can eliminate the unnecessary states within the eliminate the unnecessary states within the system.system.

The basic structure of the flow diagrams The basic structure of the flow diagrams is created in Stateflow by complex is created in Stateflow by complex if…if…then then statements which allow the program statements which allow the program to jump between two different to jump between two different algorithms .algorithms .

Solution StrategySolution Strategy

ChassisChassis

SoftwareSoftware

SensorsSensors PlatooningPlatooning StrategyStrategy

ComputationalComputationalHardwareHardware

Platooning StrategyPlatooning Strategy

The Platooning Strategy must use our The Platooning Strategy must use our sensors to allow us to follow our project sensors to allow us to follow our project constraints.constraints.

Our algorithm was developed using two Our algorithm was developed using two algorithms.algorithms.

Algorithm for Purely Algorithm for Purely Autonomous VehicleAutonomous Vehicle

M issionC onstra in ts

C artographer

P lanner

O bserver

M ap M aker

C ontro lle r

ActuatorC om m ands

P ilo t

SensorD ata

Vehic leA ltitude

D eriva tion

ActuatorFeedback

Global Map

Route Description

Local Map

Blockade Detection

Plant Segm ent

G lobal Map Updates

Algorithm for Algorithm for Vehicle Vehicle

Platooning Platooning Using a Video Using a Video

CameraCamera

(1) Edge Detection

(6a) Predict targetlocation in next

image

(7) Servo Control ofVehicle

(5) Update Motion Model

(4) Feature Aggregation

(3) Feature Measurement

(2) Data Association

(6b) Predict targetlocation at next

servo cycle

Chosen Chosen AlgorithmAlgorithm

Follow lead vehicle @1m

Follow exact path oflead vehicle

Map Maker

Planner

Observer

CMU

USSensor

Steering Control Speed Control

OtherSensors

Mission Constraints

PC104 / Simulink

Speed ActuatorSteering Actuator

SteeringEncoder

SpeedEncoder

Final Thoughts…Final Thoughts…

Questions?Questions?