1. Team XAR Autonomous Vehicle Research Group
Donald Bren School of Information and Computer Sciences

2. Student Members Team Leader: Philip Schlesinger Hussien Sleiman
Titus Sanchez
Hrayr Artunyan
Anton Popov
Lorraine Kan
Chad Christensen
Nick Mangano
Adrian Sugandhi
Titus introduces team xar and each person introduce themselves

3. Faculty Advisors Professor Crista Lopes Professor Tony Givargis
Artificial Intelligence “Driver”
Professor Tony Givargis
Embedded Systems & Robotics
Professor Isaac Scherson
Electrical Engineering
Hussien
Need picture with Professor Lopes and Scherson

4. Team Goals Develop an autonomous platform for R&D
Platform that is robust, easily controllable by software, and can serve a wide range of applications
Platform that can replace humans in hazardous environments
Platform that can complement human efforts
Lorraine
Make a platform for R&D
use platform as a substition in places where it’s bad for humans
Use platform as a compliment for rapid response scenarios

5. Accomplishments Developed a neural network obstacle avoidance system
Developed simulations
Modified electrical vehicle to become drive by wire
And most importantly….
Lorraine, Hrayr
What we’ve done to accomplish our goal(s)
Modified an electric vehicle
Neural network driving system using gps + sensor data
Sensor data from lidar
Simulations for AI-Robotics link, GPS following a modeled vehicle, control loop

6. Autonomous Driving Success
Started with simulations
First successful obstacle avoidance
GPS following and obstacle avoidance
Lorraine
Show them the videos on my laptop of the first successful run in the park (with Hrayr volunteering as an obstacle). Second video of the kewet (excellent obstacle avoidance footage!!!)

7. Next Steps
Replace hardware to increase robustness, speed, accuracy, and power
Clean up architecture and recode rapidly prototyped, software components
Fuse more sensors together
Test Improve Apply
Hussien

8. RESCUE Applications 3D Mapping Area Survey & Mapping
Transportation & Crowd Control
Nick
The above is how we are “breaking down” the stages of a rescue operation

9. 3D Mapping Build virtual representation of the world
Start with sensor data from lidars, radars, sonars, etc.
Overlay fused sensor data with live feed from cameras
Build bird’s eye view for command control training
TItus
3rd for experiencing on the ground
Bird’s eye view for a “starcraft” type view of a rescue effort.
Command control training to help lead responders coordinate better

10. Titus

11. 3D Mapping Build virtual representation of the world
Start with sensor data from lidars, radars, sonars, etc.
Overlay fused sensor data with live feed from cameras
Build bird’s eye view for command control training
Titus
3rd for experiencing on the ground
Bird’s eye view for a “starcraft” type view of a rescue effort.
Command control training to help lead responders coordinate better

12. Titus

13. 3D Mapping Build virtual representation of the world
Start with sensor data from lidars, radars, sonars, etc.
Overlay fused sensor data with live feed from cameras
Build bird’s eye view for command control training
TItus
3rd for experiencing on the ground
Bird’s eye view for a “starcraft” type view of a rescue effort.
Command control training to help lead responders coordinate better

14. Titus

15. Area Survey & Mapping Send vehicle to inspect an area Deploy sensors
Create Signal coverage map
Create Hazardous areas map
Anton
Setting a smart corridor / perimeter: Putting a bunch devices that can map out the environment and allow for wireless communication and environment information collection for quarantined areas. Detect gas.
Signal coverage map: (signal strength detection, have an intelligent algorithm decide if more scanning is needed for
Hazardous areas map: radiation, unstable buildings, spilled chemicals, fire etc)

16. Area Survey & Mapping Sample Scenario: Saving the Science Lib.
A: Send Autonomous vehicle to inspect area
B: Vehicle builds virtual reality of the surroundings and sends them back
C: Vehicle deploys an array of sensors allowing for full wireless coverage while mapping out the hazards it comes across
Anton
Setting a smart corridor / perimeter: Putting a bunch devices that can map out the environment and allow for wireless communication and environment information collection for quarantined areas. Detect gas.
Signal coverage map: (signal strength detection, have an intelligent algorithm decide if more scanning is needed for
Hazardous areas map: radiation, unstable buildings, spilled chemicals, fire etc)

17. Anton

18. Area Survey & Mapping Sample Scenario: Saving the Science Lib.
A: Send Autonomous vehicle to inspect area
B: Vehicle builds virtual reality of the surroundings and sends them back
C: Vehicle deploys an array of sensors allowing for full wireless coverage while mapping out the hazards it comes across
Anton
Setting a smart corridor / perimeter: Putting a bunch devices that can map out the environment and allow for wireless communication and environment information collection for quarantined areas. Detect gas.
Signal coverage map: (signal strength detection, have an intelligent algorithm decide if more scanning is needed for
Hazardous areas map: radiation, unstable buildings, spilled chemicals, fire etc)

19. anton

20. Area Survey & Mapping Sample Scenario: Saving the Science Lib.
A: Send Autonomous vehicle to inspect area
B: Vehicle builds virtual reality of the surroundings and sends them back
C: Vehicle deploys an array of sensors allowing for full wireless coverage while mapping out the hazards it comes across
Anton
Setting a smart corridor / perimeter: Putting a bunch devices that can map out the environment and allow for wireless communication and environment information collection for quarantined areas. Detect gas.
Signal coverage map: (signal strength detection, have an intelligent algorithm decide if more scanning is needed for
Hazardous areas map: radiation, unstable buildings, spilled chemicals, fire etc)

21. Anton
Sensor demo

22. Transportation & Crowd Control
Supply delivery
Transport disabled persons
Inform people that help is coming
Guide people to safer locations
Hrayr
Responders won’t have to worry about navigating a vehicle to deliver supplies. Supplies and people can be shuffled back and forth continuously

23. Discussion