OST-R | ITS JPO Home | Research | Research Archive

Research Archive

CICAS Workshop

December 9, 2004
Appendix A2

Breakout Session #1
Vision

Note: These notes record remarks made in six breakout groups without indicating the amount of consensus regarding the remarks either within or among the different groups.
empty cell Signal Violation Stop Sign Violation Stop Sign Assist LTAP-OD
Objective Reduce crashes due to signal violation. empty cell Provide assistance to the driver to safely move through a stop-controlled intersection (2-way) Provide information on safe maneuver at stop signs Provide assistance to drivers to safely execute a left turn. Provide gap assistance information to driver that leads to safe and effective left turns at signal intersections.
Vision in terms of CAMP model Concept 3: Infrastructure and Vehicle Cooperative - Concept 1 is on the development path (e.g. "Tool" Path); some liability concerns with deploying Concept 2
Concept 3–6 for simple signal violation; Concept 5 is overkill, but Concept 6 gives ability to warn the victim		 Concept 3 with infrastructure display: DVI and DII – to get us going with the states and then proceed to develop Concept 5 Concept 5 with deployable infrastructure-only on the evolutionary path Concept 5: deployable (i.e., post-FOT) by 2015 + Concept 1 – Concept 5 – Concept 6
With infrastructure signage/hybrid Concept 6 with V2V w/Radar
Concept 1 through modified
Concept 4 + DII (DVI by 2012)
Two-way information flows between vehicles and infrastructure.
Evolutionary path description Infrastructure-based moving toward cooperative. Will work first without the vehicle. Two-stage decision-making process/line of sight.
Deployable for:
  • Simple intersections less than 10 years.
  • Complex intersections 10-15 years.
Two alternatives for a vehicle-infrastructure cooperative system: (1) Infrastructure uses RFID to provide location references and distance information for in-vehicle processing vs. (2) DSRC with map broadcast (more capable, more expensive); need to perform trade-off study of the two alternatives. These are deployable in less than 10 years.

Can reach a "Warning Only" system by 2009. Will need to make a policy decision between the system's delivering a warning or a violation. Power will need to be at the intersection, although the RSU does not have to be at the stop sign. One RSU can transmit information for many stop signs. Assumes use of DSRC. Is a vehicle-only system possible?

RSU transmits map including the stop sign. An alternative view is for a vehicle-based system where the location of the stop bar is sent to the vehicle. Questions raised included: can the RSU support high traffic? What is the availability of intersection maps? 		empty cell DSRC included with infrastructure signage
Evolutionary path begins with Intersection Decision Support (IDS)
LTAP/OD system will vary depending on presence of stop sign or signal
State map-formed in vehicle with Concept 6, formed in infrastructure with Concept 5
Signage - both in-vehicle and in infrastructure
Vehicle capability important
Turn signal, route planned with a navigation system - need to understand driver intent
Challenges/Obstacles Pedestrians are a problem hereNeed quantifiable performance goals to make deployment decisions (safety, driver acceptance)
Increase in rear-end crashes?
By 2006, need a model-based estimate of benefits. Need benefit estimates from FOT by 2009.
Driver training? How is warning system transmitted?
Standards and validation
System deployment vs. system malfunction
How do drivers know if intersection is equipped? Or broken?
No data vs. bad data vs. broken system
Driver adaptation to system (Needs to be in the evaluation)
Overlay of CICAS information on signal data
Protected left turns vs. non-protected.
Urban/suburban intersections present bigger system challenges and bigger human factors issues. 		Relatively easy to do Driver behavior with regard to gap acceptance
Driver intent (Left, Right, Traverse)Poor intersection geometry must be fixed prior to CICAS
Expensive! Who pays?
Power to run infrastructure system components. Cost? Who pays? 		Understanding driver behavior differences
Pedestrian detection/integration in system; needs to be in state map.
Driver intent (no protected left)
Intersection geometry
Human Factors research needs for DII/DVI
Interface with traffic control systems
Utility of specified systemHard to leverage existing sensors for infrastructure sensing
Liability
Warning must be credible
Consistency of warning when delivered by both vehicle and infrastructure
Comments Will people become complacent with the capabilities of the system, in particular from the ability to warn the victim? empty cell One breakout group changed problem area title to Stop Sign Gap AssistDeployable in rural areas less than 10 years. V2I (Vehicle Characteristics transmitted to the infrastructure)This countermeasure is likely to succeed One breakout group changed problem area title to LTAP/OP for Signalized Intersections.LTAP/OD is more complex issue than 2-way stop sign gap assistNeed lots of research
Breakout Session #1
Evolutionary Path by CICAS Component
Signal Violation What do we have now in 2005? What do we want the system to look like by 2006 / the VII proof of concept demo? Are there additional technological advances that we want to target by mid-2007 for the VII field operational test? What do we think we can achieve by 2009 / pre-deployment? Comments
Sensors In-vehicle sensorsPositioning for WHICHROAD algorithmDo NOT have Concept 1 roadway, in-pavement sensorsKnow where driver is lookingKnow vehicle kinematic state, including deceleration, speed, locationHave radar, but must begin sensor research Sensors - reliable, robust, and demonstrate minimal level of accuracy in good conditions. Radar/video for Concept 1 should be ready.Driver gaze direction sensorCompleted testing to prove reliability under adverse conditions, and ability to maintain accuracy (varying environmental conditions, tracking trucks, motorcycles, etc.) Prototype WHICHLANE capability in-vehicle for FOT only (not deployable).RFID readers (range issue); non-traditional radar has to work using infrastructure alone - millimeter wave filtering vector classification One group felt that testing vehicle systems in a controlled situation - perhaps using some regular drivers - would serve testing needs in 2009.
Processing/controllers Signal controllers - can conduct an FOT now. (e.g., Signal Sniffer)
Calculations for a safe gap, but not adjusted for different drivers. 		empty cell Infrastructure processor - choose platform Advanced traffic controller capability that will probably not be commercially available (but is deployable)
Decision-making on infrastructure to provide information to all vehicles.Some decision-making on equipped vehicles.		 Research question: Do you warn both victim and violator?
Other empty cell empty cell empty cell empty cell empty cell


Stop Sign Violation What do we have now in 2005? What do we want the system to look like by 2006 / the VII proof of concept demo? Are there additional technological advances that we want to target by mid-2007 for the VII field operational test? What do we think we can achieve by 2009 / pre-deployment? Comments
Sensors Sensors on vehicles - Exp. Maps; CAN bus dataSensors on roadway - None empty cell FOT by 2007 for system alternatives proposed by one group (RFID vs. DSRC) empty cell empty cell
Communication/Data empty cell empty cell FOT by 2007 for system alternatives proposed by one group (RFID vs. DSRC). empty cell empty cell
Driver Interface empty cell empty cell FOT by 2007 for system alternatives proposed by one group (RFID vs. DSRC). empty cell empty cell
Processing/controllers empty cell empty cell FOT by 2007 for system alternatives proposed by one group (RFID vs. DSRC). empty cell empty cell
Other empty cell empty cell empty cell empty cell empty cell


Stop Sign Violation What do we have now in 2005? What do we want the system to look like by 2006 / the VII proof of concept demo? Are there additional technological advances that we want to target by mid-2007 for the VII field operational test? What do we think we can achieve by 2009 / pre-deployment? Comments
Sensors Sensors on vehicle - None
Sensors on roadway - Yes (loops)
Positioning - Radar, Lidar		 Cooperative Can run FOT in rural area (no pedestrians) with cooperative V2I communication (vehicle characteristics)
Cooperative		 Cooperative empty cell
Communication/Data empty cell empty cell Cooperative Cooperative empty cell
Driver Interface Vehicle/infrastructure Vehicle/infrastructure Vehicle/infrastructure Vehicle/infrastructure empty cell
Processing/controllers empty cell empty cell Infrastructure (basic) Cooperative (high level of data detail) empty cell
Other empty cell empty cell empty cell empty cell empty cell


LTAP/OD What do we have now in 2005? What do we want the system to look like by 2006 / the VII proof of concept demo? Are there additional technological advances that we want to target by mid-2007 for the VII field operational test? What do we think we can achieve by 2009 / pre-deployment? Comments
Sensors Sensors on vehicles - need WHICHLANE algorithm.
Sensors on infrastructure - need WHICHLANE algorithm
Need more research on alternative position systems.
GPS, DGPS at intersectionInfrastructure - Radar, Lidar,visioning
Radar (can provide presence and speed of vehicle		 Infrastructure

Infrastructure-traffic controllers that transmit current phase status		 Could get positioning from infrastructure for a FOT.

Cooperative

Distinguish type of vehicle and volume of road users - estimates

-Better pedestrian sensing,

-GPS, positioning lane-level with infrastructure correction transmitted to the vehicle		 Prototype WHICHLANE, but not deployable.

Cooperative

*Local road condition sensors
*Vehicle kinematics (including type of vehicle, speed, change in velocity, brake, turn signal, vehicle throttle pressure, etc.)
*Infrastructure - position, detects non-vehicles*Positioning, vehicle

Will not have production ready deployable system by 2009.
Communication/Data Infrastructure

DGPS data is limited. Intersection state mapGeometric Map

Range, speed, heading - V2V message set		 DSRC Frame

Phase status communicated to vehicle		 Common mini-structure message set
Warning predictive phase status		 Vehicle position, kinematics V2I
Infrastructure-signal phase information, warning.

empty cell
Driver Interface Infrastructure/Roadside
Non-MUTCD compliant DII Concepts		 Vehicle/infrastructure
In-vehicleImproved DII - test in closed course		 Vehicle/infrastructure
Downsizing for in-vehicle signage		 Vehicle/infrastructure
Integrated DVI Across safety systems (create guidelines)

Driver Behavior Issues: - What is normal behavior?
- How will drivers respond?
- Relationship of age?
- False alarm rate?
- Driving style?
- Behavioral adaptation?

Need a LOT more research.
Processing/controllers Infrastructure (loops)
Signals - look for gaps
Maps (positioning)
Experimental - Infrastructure Estimate acceleration of vehicleVehicle prototypeInfrastructure prototypeNCHRP 3-66 completed (advanced signals)		 empty cell Estimate acceleration of vehicle
Vehicle prototype
Infrastructure prototype
NCHRP 3-66 completed (advanced signals)		 Estimate acceleration of vehicle
Vehicle prototype
Infrastructure prototypeNCHRP 3-66 completed (advanced signals)

Has to be integrated with signal violation warning.
Other empty cell Software validation
Naturalistic data overlay		 Straw man approach Process for developing geometric maps of intersection

New category: Timing of warning