University of Detroit Mercy Thor

1
University of Detroit MercyThor

• Intelligent Ground Vehicle Competition, 2006

Faculty Advisors Student Team
Members Dr. Mohan Krishnan, ECE Dept. Cristian
Balas Hui-Shan Chang Dr. Mark Paulik, ECE
Dept. Kuan-Chih Chen Cheng-Lung Lee Dr. Nassif
Rayess, ME Dept. Robert McKeon Utayba
Mohammad Ricardo Moore Niyant Patel Benjamin
Radke
2
Mission Statement

• Design and build the most competitive IGVC entry
  within the scope of a short time and a limited
  budget and accounting for the abilities and
  expertise of the team members

3
Mission Statement

• Design and build the most competitive IGVC entry
  within the scope of a short time and a limited
  budget and accounting for the abilities and
  expertise of the team members

Benchmarking
4
Mission Statement

• Design and build the most competitive IGVC entry
  within the scope of a short time and a limited
  budget and accounting for the abilities and
  expertise of the team members
• Elegant and simple construction.
• Inexpensive off-the-shelf components.
• Established technology.

5
Mission Statement

• Design and build the most competitive IGVC entry
  within the scope of a short time and a limited
  budget and accounting for the abilities and
  expertise of the team members
• Dedicated, compact, purpose-built circuitry
• State of the art navigation software
• Sophisticated heuristics for decision making
• Multiple IP techniques with confidence measure

6
Overview

• Mechanical Systems
• Electrical and Electronic Systems
• Software Implementations

7
Mechanical SystemsVehicle Architecture
CATIA Animation

• Features
• Two-body articulated design.
• Front-wheel drive.
• Tank steering.
• 2 degree-of-freedom hitch.

• Construction
• Welded square tube frame.
• Fiberglass shell.

8
Mechanical SystemsDrive Train

• Features
• ½ hp DC Brushless Motors.
• Built-in encoders.
• 281 gear ratio.
• 14 pneumatic wheels.

Can be eliminated

• Construction
• Rigid, precise mounting.
• Single ½ Aluminum plate

9
Electrical SystemsPower Systems
Need work
10
Electrical and Electronic SystemsComputational
Systems

• Real Time Controller
• Integrates power regulation
• Power systems diagnostics
• Wireless Communication
• Measures vehicle motion
• Provides PWM outputs controlling motor speed
• Read real time data from IMU
• Communicate to REMOCO over Aerocomm link
• Share information with laptop

Need work
11
Electrical and Electronic SystemsSensor Systems

• Digital Camera
• LADAR ( Laser Detection and Ranging)
• DGPS (Differential Global Positioning System)
• Digital Compass/IMU (Inertial Measurement Unit)

Need work
12
Software ImplementationsImage Processing
Image Acquisition
Obstacle Removal
Adaptive Statistical Filtering
Cross-Normalized Thresholding
Edge Detection Geometric Analysis
Image Combination
Confidence Measure
13
Software ImplementationsImage Processing
Obstacle Removal
Adaptive Statistical Filtering
Cross-Normalized Thresholding
Edge Detection Geometric Analysis
Image Combination
Confidence Measure
14
Software ImplementationsImage Processing
Adaptive Statistical Filtering
Cross-Normalized Thresholding
Edge Detection Geometric Analysis
Image Combination
Confidence Measure
15
Software ImplementationsImage Processing
Cross-Normalized Thresholding
Edge Detection Geometric Analysis
Image Combination
Confidence Measure
16
Software ImplementationsImage Processing
Edge Detection Geometric Analysis
Image Combination
Confidence Measure
17
Software ImplementationsImage Processing
Image Combination
Confidence Measure (CM) CM of IP techniques
that agree
18
Software ImplementationsImage Processing
CM3, all three IP techniques agree
19
Software ImplementationsImage Processing
Simulation
20
Software ImplementationsNavigation (Autonomous
Challenge)
Algorithm 1 High Confidence Measure
Confidence Measure
Higher Speed
Fuzzy Speed Controller
Combined Steering Angle
Fuzzy Inference System
Steering Angle
IP Image
Minimum use of LADAR data
Obstacle Map
LADAR
DGPS
Compass
21
Software ImplementationsNavigation (Autonomous
Challenge)
Algorithm 1 High Confidence Measure
Confidence Measure
Fuzzy Speed Controller
Fuzzy Inference System
Steering Angle
IP Image
Algorithm 2 Low to Moderate Confidence Measure
Lower Speed
Obstacle Map
LADAR
Fuzzy Speed Controller
Combined Steering Angle
Fuzzy Inference System
DGPS
Maximum use of LADAR data
Compass
22
Software ImplementationsNavigation (Autonomous
Challenge)
Algorithm 1 High Confidence Measure
Confidence Measure
Fuzzy Speed Controller
Fuzzy Inference System
Steering Angle
IP Image
Algorithm 2 Low to Moderate Confidence Measure
Obstacle Map
LADAR
Fuzzy Speed Controller
Fuzzy Inference System
DGPS
Stored Map
Algorithm 3 Forward Motion Impossible
Low Speed
Compass
Heading
Fuzzy Speed Controller
Combined Steering Angle
Fuzzy Inference System
Turn Vehicle Around
Maximum use of all data
23
Software ImplementationsNavigation (Navigation
Challenge)
DGPS
Waypoint Map
Path Planning
Real Time Controller
Compass
Obstacle Map
Vector Field Histogram
LADAR
Schematic of VFH
GPS Outage?
IMU System
Wheel Encoders
24
Software ImplementationsNavigational Simulator
25
Vehicle Cost
26
Conclusion

• Parallel state of the art navigation algorithms.
• Multiple image processing techniques.
• Sophisticated heuristics.
• Compact, modular, clean and easy to diagnose
  circuits.
• Reliable, compact and lightweight power source.
• Good telemetry and vision systems.
• Stable, sure-footed vehicle architecture.
• Simple and reliable drive train.