Transcript CMU Kick-off Meeting - Carnegie Mellon University

General Motors Research & Development
Electrical & Controls Integration Lab
Wireless Networks for v2v and v2i
Communication
CMU Kick-off meeting
01 March 2004
General Motors Research & Development
Electrical & Controls Integration Lab
GM’s Global Telematics Vision
Connected Vehicles…Connected People
Connected vehicles save our customers time, keep them
in control, safe, informed, and entertained. When our
customers allow their connected vehicles to be part of a
network, it makes everyone’s life better by generating
information that benefits everyone.
ECI Focus:
Explore the feasibility, limitations and potentials of wireless
networks to deliver real-time traffic information services to
our customers and to demonstrate automotive safety
applications for GM vehicles
01 Mar. 2004
2
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Ubiquity
Network Topology
Wireless Networks
• Mobile Ad-hoc (Self-forming, multihop)
• GSM, CDMA, PCS, 2.5/3G
• Wi-Fi (802.11x)
• XM Satellite
Traffic
Probe
Vehicle
01 Mar. 2004
3
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Example Traffic
Scenario
Exit
• Immediate spread of knowledge to
surrounding vehicles within ad-hoc
network
• Incidence notification to OnStar
• Information relay to other vehicles for
dynamic route guidance
01 Mar. 2004
4
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Objectives
Safety:
• Develop a cooperative collision warning strategy
using V2V communication
• Replace/enhance collision avoidance sensors for
GM vehicles
Telematics:
• Enable advanced telematics services through
network
• Create OnStar to the power of N
01 Mar. 2004
5
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Our Approach
• Explore mobile ad hoc network concept using
proprietary and standards based solutions
• Develop and demonstrate active safety and
possible telematics applications to assess
feasibility and establish requirements
• Conduct research to address network
performance, scalability and capacity using
analytical and simulation tools for real-world
scenarios
• Leverage expertise of our research partners and
integrate the concept into GM vehicles
01 Mar. 2004
6
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Mobile Ad Hoc Network Solutions
•
Proprietary Solution from Mesh Networks
–
–
–
–
•
2.4 GHz unlicensed band
Off-the-shelf hardware
Proprietary routing and communication protocols
Ready for quick integration and concept validation
Standards - DSRC
–
–
–
–
–
01 Mar. 2004
5.9 GHz licensed band
FCC approval granted in Dec. 03
Communication protocols under development
No commercial H/W available yet
No plans for network routing protocols
7
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Our Plan
•
By Fall ’04:
–
–
Develop and demonstrate Rear-End, Intersection, Lane
Change/Merge countermeasures and collect real-time vehicle
data using commercially available hardware
Conduct ad hoc network research to answer:
•
–
Conduct network simulation research to answer:
•
•
Network performance, RF power control, network scalability, etc.
RF propagation model, data channel capacity, effect of vehicle
density, etc.
By Fall ’05:
–
–
01 Mar. 2004
Integrate research results into concept vehicles
Investigate possibility to enhance or replace DAS (Driver
Assistance System) for production
8
Jay Parikh
ECI Lab
General Motors Research & Development
Safety
Applications
Normal Driving
 Cooperative Adaptive Cruise Control
 In-Vehicle Signage
 In-Vehicle Amber Alert
Safety Warnings
Cooperative V-V Collison Avoidance
 Emergency Electronic Brake Lights
 Cooperative Forward Collision Warning
 Blind Spot Warning
 Lane Change Warning
 Vehicle-to-Vehicle Road Condition Warning
 Vehicle-to-Vehicle Road Feature Notification
 Visibility Enhancer
 Highway Merge Assistant
 Cooperative Collision Warning
 Approaching Emergency Vehicle Warning
01 Mar. 2004
Electrical & Controls Integration Lab
Safety Warnings
Intersection Collision Avoidance (Infrastructure-Assisted)
 Traffic Signal Violation Warning
 Stop Sign Violation Warning
 Left Turn Assistant
 Stop Sign Movement Assistant
 Intersection Collision Warning
 Pedestrian Crossing Information at Intersection
 Emergency Vehicle Signal Preemption
Non-Intersection Collision Avoidance (InfrastructureAssisted)
 Curve Speed Warning – Rollover Warning
 Low Bridge Warning
 Low Parking Structure Warning
 Work Zone Warning
 Wrong Way Driver Warning
 Road Condition Warning
 Blind Merge Warning
 Highway/Rail Collision Warning
Collision Mitigation
 Pre-Crash Sensing for Cooperative Collision Mitigation,
e.g. enhanced air-bags & seat-belts, truck/car compatibility,
brake assist
Post Collision and Other Safety
 Post-Crash Warning, e.g. ACN
 SOS Services
9
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Example Application Scenarios
Avoiding lane change
collision
Vehicle brakes hard
Collision mitigation
Avoiding rear-end
collision
V-V messages
Avoiding intersection
collision
Traffic signal
01 Mar. 2004
10
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Vehicle as Traffic Probe
• Vehicles periodically report its speed and position to a
data center for real-time traffic
• Vehicles directly communicate with data center
• Data center must be capable of managing
communication channels and data from large number of
vehicles
• How can ad hoc network help?
– Reduce communication requirements by aggregating and
processing data from networked vehicles before reporting to data
center
– Reduce required data transfer
– Reduce dependability on infrastructure
– Quickly spread emergency information among networked
vehicles
01 Mar. 2004
11
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
UC Berkeley / PATH
• Demonstrate active safety applications - Fall ‘04
– Utilize 6 AHS vehicles (Buick) to demonstrate:
• Rear-end crash warning
• Intersection warning
• Lane change/merge warning
– Develop protocols and algorithms for cooperative situational
awareness
• Establish communication content and performance requirements
• Develop and demonstrate warning / interface strategy
• Collect, analyze, and visualize data
– Implement using commercially available hardware solution from
Mesh Networks
01 Mar. 2004
12
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Carnegie Mellon University
• Create and demonstrate ad hoc network based on
evolving DSRC standard capable of supporting safety
and telematics applications – Fall ‘04/05
– Define latency requirements for:
• Network connection and packet transmission/retransmission
– Determine network scalability and reliability
• How many hops can ad hoc network reliably cover?
– Optimize protocols for automotive environment
• Routing protocols - proactive, reactive and hybrid
• Targeting or controlling a flood fill
• Adaptive power control for sparse/dense traffic
– Develop algorithms to cooperate between Mobile IP, ad hoc
network and infrastructure
– Demonstrate real-time vehicle data collection using ad hoc
network and communication with infrastructure
01 Mar. 2004
13
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Hughes Research Laboratory
• Develop and validate a cooperative 360° collision
warning strategy through simulation - Dec ‘04
–
–
–
–
–
–
–
Develop RF propagation model
Simulate communications protocols
Evaluate channel capacity
Simulate network throughput
Simulate dense and sparse traffic conditions
Simulate ranging schemes
Integrate with CORSIM traffic simulator for real-world
traffic scenarios and validation
– Evaluate possible impact on traffic flow and vehicle
safety
01 Mar. 2004
14
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Ad Hoc Network for Active Safety and Telematics
R&D Approach:
• Explore ad hoc network using commercial and
standards based solutions
• Develop and demonstrate active safety and
telematics applications
• Research to address network performance,
capacity and scalability using analytical and
simulation tools for real-world scenarios
• Leverage expertise of our collaborative research
partners and integrate the concept into GM
vehicles
CMU:
• Demonstrate based on DSRC – Fall ’04/05
• Develop a test-bed using 5.8GHz 802.11a for:
• Real-time propagation model
• RF channels management (control and data)
• Determine network scalability, reliability, latency
• Develop and optimize routing protocols
• Develop algorithms to cooperate between ad hoc
network & infrastructure for telematics applications
Budget – CRL $$$
Resources – 2 Faculty, 1 post doc, 4 grad students
01 Mar. 2004
UC Berkeley / PATH:
• Demonstrate active safety applications – Fall ’04
• Forward collision warning
• Intersection warning
• Lane change/merge warning
• Develop protocols and algorithms for situational
awareness
• Implement using commercially available solution
Budget – 460k
Resources – 1 Faculty, 2 Res. Eng., 1 Grad Student
HRL:
• Develop and validate a cooperative 360° collision
warning strategy via simulation – Dec. ’04
• Develop RF propagation model
• Simulate communication protocols
• Simulate different traffic conditions
• Simulate network throughput
• Evaluate impact on traffic flow and vehicle safety
Budget – 275K
Resources – ¾ FTE
15
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Moving Vehicle Results
• 1 stationary, 1 dynamic antenna, with obstruction
• Receiving vehicle ~ 40 mph, no lost packets
dy1
v1
f
dx1
01 Mar. 2004
16
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Moving Vehicle Results
• 2 vehicles ~ various speeds, same direction
• Distances up to 150 m, no lost packets
v2
v1
dx1
dy1
x
z
y
01 Mar. 2004
17
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Moving Vehicle Results
• Various vehicle speeds, same direction, 150 m
• Cut-in SUV, no lost packets
v2
v1
dx1
dy1
01 Mar. 2004
18
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Moving Vehicle Results
• Highway, light traffic, 60mph
• Reception good, generally dependent on line of sight
large truck blocking
v2
v1
dx1
dy1
01 Mar. 2004
19
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Internal Project Plans for 2004
• Investigate V-V communications and GPS as a
low-cost solution for remote sensing between
vehicles
• Development of a 360 degree collision warning
strategy using V-V communications
• Analyze, develop and demonstrate prototype
vehicle safety applications using V-V
communications
01 Mar. 2004
20
Jay Parikh
ECI Lab
General Motors Research & Development
Electrical & Controls Integration Lab
Issues and Challenges
•
•
•
•
•
•
•
•
•
•
•
Analysis of communication technologies and standards
Network protocols for V-V communication
Standardized data messages
Communication channel capacity and availability
Infrastructure integration
Range of coverage, intelligent power management
Interference, connection reliability
Connection Latency
Security, Privacy
(D)GPS/Map integration
Antenna, etc.
01 Mar. 2004
21
Jay Parikh
ECI Lab