LOS Throughput Measurements in Real-Time with a 128-Antenna Massive MIMO Testbed
CoopPowerPoint_7_30
1. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 1
Bi-Directional Cooperative
Obstacle Avoidance for
Heterogeneous Unmanned
Vehicles
By: Jonathan Lwowski
The University of Texas at San Antonio
Department of Electrical and Computer Engineering
Advisor: Dr. Daniel Pack
B.S. Computer Engineering
2. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 2
Overview
• Motivation
• Significance of Methods
• Obstacle Avoidance Algorithm
• Door Entering Algorithm
• Simulation Results
• Conclusion
• Future Work
• REEF Projects
3. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 3
Motivation
• Reactive Bearing
Angle Only Obstacle
Avoidance System for
Unmanned Ground
Vehicles
• Global Vs. Local Path
Planning
• Visual Electro-Optic
Sensors
4. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 4
Significance of Methods
• First reported bi-directional heterogeneous system
using only cameras.
• Allows the UGV to incorporate obstacle avoidance
into global path plan
5. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 5
Flow Diagram
6. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 6
Obstacle Avoidance Algorithm
• Doorway is not detected
• UGV Tracking
• Directional Path Calculation
• Criteria for Ending Obstacle Avoidance Procedure
• Communication to UGV
• UGV Controller
7. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 7
UGV Tracking
• Color Thresholding
• PD Controllers
8. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 8
Directional Path Calculation
• Detect Obstacles
• Create Nodes
9. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 9
Directional Path Calculation
• Calculate Shortest Path Using Dijkstra's Algorithm
• Convert Path to Directional Information
10. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 10
Criteria for Ending Obstacle Avoidance
Procedure
• Determine if UGV has passed the obstacle
• Acute Angle
• Obstacle Passed
• Obtuse Angle
• Obstacle Not Passed
11. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 11
Communication to UGV
• Direction Sent Check
• Directional Path
• Obstacle Passed
12. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 12
UGV Controller
• Uses the bearing angle as
the only input
• Directional path determines
which direction to avoid the
obstacle
13. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 13
Simulation Results
14. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 14
Doorway Entering Algorithm
• Doorway is detected
• Door Entering Using PD Controller
• MAV and Doorway Tracking
• Homography Decomposition
15. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 15
Door Entering Using PD Controller
• PD Controller
• Centroid of Doorway
16. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 16
Simulation Results
17. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 17
MAV and Doorway Tracking
• Create four feature
points on door and MAV
• Find the centroid of
each feature point in the
reference frame
• Find the centroid of
each feature point in the
current frame
18. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 18
Homography Decomposition
• Calculate homography matrix using feature points
• Four feature points from reference frame
• Four feature points from current frame
• Decompose homography matrix to find camera
rotation and translation matrix
• Homography Matrix
• Camera Calibration Matrix
19. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 19
Conclusion
• First reported bi-directional heterogeneous system
using only cameras.
• Allows the UGV to plan a long term path
• No requirement of GPS signals
• Effectiveness of system demonstrated using
software simulations
20. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 20
Future Work
• Finish homography based door entering
• Implement system on to hardware
21. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 21
REEF Projects
• Topological Map Merging
• Calculate Elevation Map
22. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 22
REEF Projects
• Calculate Elevation Map using 20 Loops
• Calculate Elevation Map using 200 Loops
• Combine Elevation Maps
23. The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249 23
Acknowledgments
• AFRL Munitions Directorate
• OSD & The Privileged Sensing ARPI
• Mathematical Modeling and Optimization Institute
• Dr. Emily Doucette
• Dr. Will Curtis
Questions?