2016 D-STOP Symposium ("Smart Cities") session by SwRI's Paul Avery. Get symposium details: http://ctr.utexas.edu/research/d-stop/education/annual-symposium/

1. Cooperative Systems R&D
Southwest Research Institute
IVS.SwRI.ORG
Connected and
Autonomous
Vehicle Systems
R&D Overview
Paul A. Avery
Principal Engineer – Cooperative Systems
Southwest Research Institute
+1.210.522.6732
pavery@swri.org

2. 2
Let’s Talk About…
 Connected Vehicles
 Automated Vehicles
 Cooperative Systems
Slate.com

3. Connected Vehicles
Illustrationsource.com

4. 4
En.wikibooks.org
Connected Vehicles Communication
Illustrationsource.com
However!
Special Applications Like Platooning

5. 5
What is A Connected Vehicle?
Concentric Yellow Circles?
Lines and Dots?

6. 6

7. 7
zzzzzzzz……..

8. 8

9. 9

10. V2I – Overheight Detection,
Notification, and Autonomous Disable

11. 11
V2V – Blind Spot Warning

12. V2I – Railroad Grade Crossing Notification

13. Environmental Efforts:
Integration into a Traffic Management System
• Florida’s SunGuide
enhanced to include
environmental data
• Average emissions
• Total emissions
calculated from traffic
volume
• Number of samples
• Data aggregated by
roadway segments
• Only requires a sample of
emissions data to
calculate overall
emissions impact

14. Ooops….

15. Vehicles are More Complex…
(and Connected Vehicle is only part of it)
 1965:
 No computers
 No software
 2015:
 Up to ~200 computers
• Consider TPMS are 4 computers and wireless…
 >100 million lines of code
 LTE (or similar) enabled vehicles are becoming commonplace
http://www.informationisbeautiful.net

16. Challenges with Connected Vehicles
 Recent attacks on Connected
Vehicles:
 Jeep Cherokee: “Hackers Remotely
Kill a Jeep on the Highway—With
Me in It”
 GM OnStar: “This Gadget Hacks
GM Cars to Locate, Unlock, and
Start Them”
 Tesla Model S: “Researchers
Hacked a Model S, But Tesla’s
Already Released a Patch”
 Impact of these attacks:
– Erodes public trust
– Raises awareness – improves security
practices
– Not a setback for DSRC
Image Source: Wired Magazine

17. Automated Vehicles

18. 18
http://william-flew.com/autocars/problems-not.html
Future of Traffic Enforcement?

19. 19
Source: All over the internet….
License and Source Code Please…

20. 20
!Google_Car

21. 21
Interacting with Humans….
Displays a cross walk for the
human
Lights “move” to indicate
direction in which human
should proceed
Fastcodesign.com

22. 22
Maybe Releasing Autopilot Into the Wild
Wasn’t A Great Idea?

23. 23
SwRI’s
Automated
Vehicle
Technology
Development
Off-RoadOn-Road
• 8 Fully-Automated
Different Platforms
• On- and Off-Road
Navigation
• Single- and Multi-Modal
Low Cost Sensor
Configurations
• Not dependent on GPS
• Safe Operation near
Vulnerable Road Users
• Platform, hardware, and
sensor agnostic.

24. 24
But How Do We Avoid This?

25. Connected-Automated Vehicles
Worldreviews.com
(But Only If The System Is Cooperative)

26. 26
Its.dot.gov
Vehicle Platooning
• Rules must be
established for joining a
team: Max/min speed,
following distance, vehicle
performance/efficiency
characteristics, etc
• Special message must be
used, including a “Team
ID” to enable cooperation
among specific vehicles

27. Cooperative Truck Teaming with Sensor Sharing
V2V Only SimulatedSimulated
Sensor Horizons
 Commercial trucks form a
cooperative team
 Share data regarding other vehicles
 Enables enhanced situational
awareness for all vehicles

28. 28
Intersections Too

29. 29

30. 30

31.  Developed under SwRI IR&D program
Mobile Autonomous Robotics Technology Initiative
(MARTI)
 Cooperative vehicle behaviors for sharing sensor
data in a dynamic traffic environment
 Utilizes the sensing capabilities of other vehicles
as extension of perception system
 Use-case: Pedestrian safety
 Communicates V2V regarding position, speed,
and heading of pedestrians at intersections
 Receiving vehicle determines appropriate
action
 Jointly developed by SwRI (US) and INRIA
(France)
 Patent 7,994,902 awarded August, 2011
 Jointly demonstrated in 2008
 Versailles, France
 New York City, NY (ITS World Congress)
Cooperative Sensor Sharing for
Unmanned Systems
SwRI’s MARTI in Manhattan 2008
Cooperative Sensor Sharing
SwRI/INRIA in Versailles France 2008
Cooperative Sensor Sharing

32. Cooperative Vehicle Lane-level
Model Generation Using DSRC
 Vehicles broadcast message on
their location (Basic Safety
Message)
 Roadside device collects
messages for processing
 Using learning algorithms, the
noisy data is reduced to lane-
level GPS coordinates
 Data is reduced to minimum
required set, and then re-
broadcast to vehicles
 If obstruction is removed,
vehicles will begin traversing
the area again, and a new map
will be processed and re-
broadcast

33. 33
Conclusions?
Connected vehicles are able to send and receive
information about themselves, other vehicles, and their
environment.
Automated vehicles are able to sense and respond to their
environment within the constraints of their component
technologies and programming.
Neither technology by itself provides a comprehensive
solution to our traffic challenges.
The technologies combined have the potential to
significantly enhance and improve many aspects of
society.

34. Cooperative Systems R&D
Southwest Research Institute
Thank You