Meeting Summary - August 27, 2010

Produced by:
Volpe National Tranportation Systems Center
Research and Innovative Technology Administration
U.S. Department of Transportation

Executive Summary

IntelliDrive Safety Workshop

The Intelligent Transportation Systems (ITS) Joint Program Office (JPO) hosted a three day connected vehicle workshop from July 20th – 22nd, 2010 in Chicago, Illinois. The objectives of the workshop were to provide an informational briefing on the status of connected vehicle Safety Programs, solicit stakeholder feedback on the program roadmaps, and discuss policy issues critical to supporting deployment. On the first day of the workshop, presentations were given on the technical roadmaps for each connected vehicle Safety research program. Stakeholders were encouraged to provide input on program roadmaps and ask questions. The second day of the workshop included a bidder’s conference for DSRC devices and an overview of the connected vehicle Policy program. Finally, the last day of the workshop was dedicated to smaller breakout sessions focusing on the specific topics of cost benefits analysis and the Vehicle-to-Infrastructure program.

Introductory Presentations

The first day of the workshop was devoted to brief overviews of key topics in the connected vehicle Safety program.  The presentation topics included the following:

• Connected Vehicle Safety Program Overview, Mike Schagrin, [PDF1MB]
• Vehicle-to-Vehicle and Safety Pilot, Ray Resendes [PDF 782KB]
• Vehicle Safety Communications Working Toward V2V Deployment, John Harding [PDF 192MB]
• Commercial Vehicle V2V Safety Program, Alrik Svenson [PDF 120MB]
• Human Factors for Connected Vehicle, Stephenie Binder [PDF 785KB] 
• Connected Vehicle Safety Program Certification Overview, Walton Fehr [PDF 772KB]
• Vehicle to Infrastructure Connectivity for Safety Applications, Monique Evans [PDF 229KB]

List of Presenters

Mike Schagrin
Program Manager, Safety
ITS Joint Program Office
Research and Innovative Technology Administration
U.S. Department of Transportation

Ray Resendes
Chief, Intelligent Technologies Research Division
National Highway Transportation Safety Administration
U.S. Department of Transportation

John Harding
Research Engineer, Intelligent Technologies Research Division
National Highway Transportation Safety Administration
U.S. Department of Transportation

Alrik Svenson
Research Engineer, Office of Applied Vehicle Safety Research
National Highway Traffic Safety Administration
U.S. Department of Transportation

Stephanie Binder
Human Factors Engineer
National Highway Traffic Safety Administration
U.S. Department of Transportation

Walton Fehr
Program Manager, Systems Engineering
ITS Joint Program Office
Research and Innovative Technology Administration
U.S. Department of Transportation

Monique Evans
Director, FHWA Office of Safety RD&T
U.S. Department of Transportation

Connected Vehicle Safety Program Overview
Mike Schagrin

Presentation Outline

• Connected Vehicle Program Structure
• Major Connected Vehicle Objectives
• Evolution of Connected Vehicle Deployment
• Opportunity for Safer Driving
• Key Elements of the Connected Vehicle Safety Program
• Connected Vehicle Safety Program Areas
• Leveraging Safety Program Activities to Provide Cross-Cutting Support
• Technical and Policy Interactions
• Outstanding Technical Issues
• Outstanding Safety Policy Issues
• Large Scale Multi-Application Pilot

Key Comments

No comments.

Questions & Answers

No questions.

Connected Vehicle-to-Vehicle and Safety Pilot
Ray Resendes

Presentation Outline

• Who Is NHTSA / NHTSA Congressional Authority
• Rulemaking Requirements and Requirements for FMVSS
• Sources of Rulemaking Action and Rulemaking Process Overview
• Vehicle Safety Communications
• V2V Safety Application Research Plan and Assumptions
• V2V Program
• Safety Pilot & Safety Pilot Roadmap
• Track 1: Light Vehicle Driver Clinics & Performance Tests
• Track 2: Model Deployment

Key Comments

• The perspective of the discussion seems strongly US-centric.  The existing avenues for international cooperation seem relatively weak. There should be more direct cooperation.

Questions and Answers

• How long will it take from the 2013 decision before we see a NPRM?
  • Approximately 12 months.
• What international harmonization will take place?
  • There is a bilateral agreement with the European commission, and we are in the process of identifying standards and regulations that need to be harmonized. Our intent is to get vehicle manufacturers to help us identify those standards.
• Will NPRM and rulemaking be strictly a NHTSA matter or will the FCC be involved?
  • The JPO is taking the lead in working with the FCC.
• What is the difference between an Integrated Vehicle and a Cooperative Vehicle?
  • An integrated vehicle incorporates a full suite of safety applications, while a cooperative vehicle only broadcasts a “Here I Am” message.
• From a software perspective, how many test cases will be needed to get certified?
  • Typically, as objective test criteria for crashes, we would need a handful of scenarios. Functional safety is the vehicle manufacturer’s responsibility so they have to do their own number of tests. There is still a question of whether the government needs to get involved in the functional safety testing area.
• Where does the large scale multi-application pilot happen on the schedule and how does the completion of it interact with the decision point being discussed?
  • The large scale pilot is a potential project currently under discussion at US DOT.  It will be primarily a V2I pilot that includes safety, mobility, and environmental elements, as opposed to the safety pilot which is primarily safety.  The timeframe is 2-3 years after the safety pilot is completed.  It will have no influence on the 2013 NHTSA decision.
• Wouldn’t it be better if there were more fully integrated vehicles?
  • There will be 60 fully integrated vehicles and 2,000 to 3,000 cooperative vehicles- we believe this is sufficient to collect the right amount of data. There are also funding constraints.
• Could you please elaborate a bit more about the nomadic devices?
  • We are looking just at “Here I Am” devices and broadcasting the basic safety message. There is a concern that additional technology applications might cause more driver distraction.
• You mentioned your intent to cover multiple vehicle platforms—how do other commercial vehicles fit into schedule?
  • The agency hasn’t committed to a regulatory decision at that point.  The safety pilot will include trucks.
• Don’t you need absolute positioning capability to avoid guardrails, bridges, and other static objects?
  • We eventually do see this capability being added.
• Do you envision any infrastructure at all in the Safety Pilot?
  • We will have an implementation of a security solution which may require infrastructure support. We will have roadside equipment at intersections broadcasting signal phase and timing (SPaT) information to see if it impacts the ability of V2V to operate.
•  The European Union is doing a lot of large scale trials. How much cooperation is really going on with the EU? Is it just monitoring?
  • There are ongoing discussions between the European Union and the United States to see what each other’s research plans are and to see what can be done to colloborate.
• What will be evaluated in the model deployment?  Six months doesn’t seem like enough exposure.
  • The model deployment is based on the forward collision application and using other field tests. We looked at the number of miles traveled and crashes and have done statistical analysis to show that 60 vehicles and 2000 cooperative vehicles will produce enough information to get a valid number of samples. We need to weigh the cost of additional vehicles.
• Are there any specific harmonization forums that are underway?  The view in here seems to be US-based.  Are there any plans to encourage direct global cooperation?
  • From the standards perspective, we have reached an agreement with the European Commission on harmonization.  A forum has not yet been decided.  The process for information exchange is very slow.
• Has anything been specified over who might participate in the Safety Pilot?
  • The Safety Pilot will be in the US.  We have not made any restrictions on the team yet but we may establish some.  We expect and hope that as teams form they will be multi-faceted in their approaches toward trying to achieve the requirements.  There will be a bidders’ conference as well.
• Are we looking at a doubly rare event (high risk and close proximity)? Was this taken into account in statistical simulations?

• Yes, that is what drove the number of cooperative vehicles.

Connected Vehicle Safety: Communications Working Toward V2V Deployment
John Harding

Presentation Outline

• Build V2V—Technology and Policy
• Track 1—Crash Scenario Framework
• Track 2—Interoperability
• Track 3—Benefits Assessment
• Track 4—Application Development
• Track 5—Driver Issues
• Track 6—Connected Vehicle Policy Issues
• V2V Traditional Aspects
• Requirements for FMVSS
• Building V2V—Compliance
• V2V Non-Traditional Aspects
• Building V2V—Operations, Aftermarket Devices, and Business Models

Key Comments

• The different tracks seem to be highly interdependent. Clearly presenting the linkages and overlapping activities would be helpful.

Questions & Answers

• How do you reconcile a business model with a mandate? What does NHTSA think about this? Where are the opportunities?
  • From our standpoint it creates opportunity because businesses will know a level playing field exists.
• As we go through the tracks, there are lots of linkages.  Is there a version that shows linkages?
  • Yes, there are lots of interdependencies.  While programs were crafted more or less on their own, we are going through an activity to state linkages more explicitly.  This will be available in a few months.
• V2V applications are fundamentally no different from a driver’s perspective than if they are done with sensors.  Why is this part of the connected vehicle program as it’s a much bigger problem for automotive warning systems in general?
  • We are dealing with this in our human factors work. Here we are looking at it from a safety applications sense and what additional performance capability can be achieved. There are nuances that DSRC has over radar and there are vehicle control possibilities. We are looking at the DVI and driver reactions.
• Somebody will know when crashes are avoided.  Is there any consideration that information will be gathered on how effective the system is?
  • Yes, the statistical analysis group is looking into how to monitor performance in to the future.
• The language on reliability and fault-tolerance seems to be missing.  What are the failure modes?  Does compliance include a standard for reliability?
  • These are the questions that we are trying to answer.  We are looking at criteria for identifying a fault and at how to inform the driver or the Certificate Authority.
• How will you incorporate reliability and a systems engineering approach?
  • We will attempt to address this in compliance standards.
• How are standards for liability being addressed?
  • We are looking at the connection between the policy and technology side. 
• Interoperability is a key issue.  It is highly likely in the US that we will not be finished by the time that Europe and Asia have adopted their guidelines. Your thoughts?
  • We have different environments but we will attempt to coordinate where it makes sense.  We realize that they are working faster and are attempting to ensure collaboration.
• Do you see a broader role of “Here I am” devices other than just being an environment for others?
  • There is a focus on safety for now but at some point in the evolution we may incorporate additional functionality.

Connected Vehicle Commercial Vehicle V2V Safety Program
Alrik Svenson

Presentation Outline

• Connected Vehicle for Commercial Vehicles
• Track 7: Commercial Vehicles
• Commercial Vehicle V2V Safety Application Research Plan
• 2010 Commercial Vehicle V2V Research
• CV Interoperability Issues
• Performance Requirements
• Driver Vehicle Interface (DVI) Specifications
• CV Research Timeline

Key Comments

No comments.

Questions & Answers

• To what extent does commercial vehicle V2V mean talking to other commercial vehicles vs. all other vehicles?
  • The program is looking at all aspects.
• At what point in time will you start involving truck OEMs in the process?
  • We have already involved OEMs and suppliers.  UMTRI is coordinating with activities of CAMP.
• How will the driver interface be different for Commercial Vehicles than for light vehicles?
  • The layout of vehicle is different.  Commercial drivers have accountability requirements that are different.  The vehicles are larger and have different handling characteristics.  Drivers have additional training and are also dealing with additional data such as dispatchers.
• What about rural environments?
  • Research in the CV area can apply to all areas.  All of our work is applicable for any environment or vehicle.
• There is no mention of ATA.  Are they involved?
  • Yes, they were involved in the CVSA San Antonio workshop held in April 2010.
• Are emergency vehicles included in CV work?
  • The initial research is applicable to any heavy vehicle including emergency vehicles.

Human Factors for Connected Vehicle
Stephanie Binder

Presentation Outline

• What is Human Factors?
• Human Factors for Connected Vehicle
• Human Factors for Connected Vehicle: Research Plan
• Track 1: Problem Definition
• Track 2: Performance Metrics
• Track 3: Integration Strategy
• Track 4: Long-term exposure
• Track 5: Outreach
• V2V Plan: Tracks 5 and 7
• NHTSA Distraction Plan
• Primary Outcomes

Key Comments

No comments.

Questions & Answers

No questions.

Connected Vehicle Safety Program Certification Overview
Walton Fehr

Presentation Outline

• Connected Vehicle Program Structure
• Major Connected Vehicle Objectives
• Certification Task Objectives
• Example: V-V Crash Avoidance
• Certification Framework
• Test the framework/Provide a service

Key Comments

No comments.

Questions & Answers

• In the V2V example, the transponder is shown sending DSRC messages.  Is this an error?
  • This is a “what if” scenario.  DSRC tags are seen as potentially multipurpose and could be used for transactional purposes.
• Has any thought been given to stress testing?
  • This is getting away from certification issues and entering the area of validation.  When we define the interface and create performance requirements someone will need to validate those requirements.
• Are there specifications available that can be crafted to talk about minimum transmission range as an example of integration level testing?
  • We’ve taken our best attempt to craft the requirements that have been issued for the “Here I am” device solicitation. Seven months of time is included to determine if the specifications are sufficient for this purpose.
• What is process for getting actual testing specification stated, and who will provide it?
  • This is one of the first tasks—to generate a requirements document that articulates the types of tests needed.
• Is there not a 4th level—vehicular level evaluation in the presence of other vehicles attempting to do the same thing?
  • Impractical to do during the safety pilot time frame, but ultimately, yes, if deemed by the stakeholder community to be necessary this type of testing would be included.
• I’m seeing lots of emphasis on HIA devices, but I see this as the first opportunity to test full devices.  I hope the focus on transmitting devices only is a false one.
  • We are using this as a service to HIA device makers, but all devices will need to be certified to some standard.  Please don’t confuse this HIA certification effort with the full device certification effort.
• I presume from your certification presentation that there needs to be some sort of conformance activity prior to the certification activity.  How do we address certification on the platform, i.e. automotive environment?
  • We need to sidestep that complexity at this stage.  “Here I am” devices are transmit-only so we don’t need to worry about additional levels of complexity on the back end of the interface.  Complete certification of the interface will be much more complex than what is being done here.
• Do you see these tests as being simple pass/fail, or levels of performance?
  • That depends on the results.  If lots pass, the criteria will be pass/fail.  If, for example, none pass, then performance standards will be re-examined.

Connected Vehicle to Infrastructure Connectivity for Safety Applications
Monique Evans

Presentation Outline

• Connected Vehicle Program Structure
• Safer Driving in the V2I Environment
• V2I Safety Focus Areas
• V2I Safety Research Plan
• V2I Safety Applications Roadmap
• Track 1: Applications Analyses
• Track 2: Prototype Development
• Track 3: Infrastructure Communications & Interoperability
• Track 4: Benefits Assessments
• Track 5: Deployment Planning
• Breakout Questions

Key Comments

• There seems to be ‘stove piping’ between V2V and V2I. It is difficult to see where these things come together and where decisions get made.  From a policy perspective the partitioning makes sense.  But from an engineering side there are opportunities for shared data and collaboration.
• Adaptive signal control should be included in V2I testing.
• There are a number of safety issues that have nothing to do with vehicle communications. There is concern that the DOT will only focus its attention on communications for safety.  
• Security has been mentioned a number of times, including certificate authority.  We really need concrete definitions of what the CA equipment will look and feel like.

Questions & Answers

• It seems that the Field Operational Test runs significantly beyond the decision point in 3Q 2013.
  • The decision point for 2013 is really related to V2V.  V2V is looking at rulemaking, and V2I is looking at guidelines, so they don’t coincide on timing.
• Many believe a number of applications could be solved either way.  Isn’t there a need to pick the best course?
  • V2V side is rulemaking; V2I side is more issuing policy guidance.  Alignment of roadmaps is an activity currently being undertaken at DOT.  FHWA does have rulemaking authority. V2I and V2V address a different set of problems.
• It looked like NHTSA was going to make decision purely based on V2V results.  It also looked like analysis shown here for V2I isn’t going to be complete. It appears that the hard 2013 rulemaking will not include some of this data.
  • That’s correct.
• As a comment, inevitably there will be crossover.  As a question: are grade crossings included in V2I program?
  • We are not excluding any intersection types.  We have had discussions with FRA. However, those types of crashes (rail crossings) are a very small percentage.  We’d like to find ways of representing these types of crashes that are not well represented but still impact safety.
• What exactly are you going to use the intersection DSRC units for in the Safety Pilot?
  • Intersection DSRC units will be used exclusively to transmit SPaT info.  We do not anticipate receiving data for real time intersection control.
• The roadmap seems centered around traffic signals.  (SPAT, CICAS SLTA, etc).  MN has already implemented “stop sign assist” using DSRC.  I don’t see any of this work shown here.
  • We were trying to keep this high level; we’re not ruling anything out. Rural crashes account for many crash related injuries. We’re not ruling out this type of intersection.  These may have safety benefits, but we don’t know exactly what the benefits are from implementing V2I in these locations yet.  It’s also not clear that all of these instances are served better by V2I than by a V2V solution.
• If there are non-federal agencies/companies that want to deploy something faster than roadmaps, does the agency have a policy on this yet?
  • US DOT doesn’t currently have a position.  We’re trying to be open and transparent regarding our plans—that’s what these roadmaps are about.  If a company wants to deploy something that’s up to them, but we want to make sure they’re aware of what else is going on and can decide whether it makes sense to deploy something when certain technical issues remain unresolved.
• Where and when will Safety Pilot be deployed?  Multiple locations?
  • There will be two stages.  The first stage (driver clinics) will be in 6 locations, and the second stage (model deployment) will be in one location, with location TBD.  A big element of the bidding process will look at how a team can structure a model deployment that meets our requirements.  There are no predefined locations.
• In last June’s workshop it was recommended that a backhaul network be described.  Any additional action on this?
  • We’re looking at it, but security is still a big issue.
• When will the detailed specs on the CA be issued?  Where is this in the roadmap?
  • The requirements are currently being defined by CAMP.  We don’t know the frequency with which certificates need to be issued.  These types of questions will drive the specs of the CAs themselves. Also, the Policy /Technical tradeoffs need to be addressed first.  Both the CAMP and systems engineering activities currently underway will help address these issues.
• We’ve heard lots about the singular safety pilot.  It would seem that you would want multiple teams implementing these in multiple locations to learn more.  Is this a budgetary constraint?
  • You’re right—different environments have different considerations.  For driver clinics this is why we’re using approximately six locations.  The model deployment is very expensive.  We feel we’re better off with a robust singular environment rather than multiple lesser environments.
• I saw a number of schedules presented.  This now being 3Q 2010, we’re roughly a third of the way through the five year plan.  How are we doing?  Are we on schedule? 
  • I think we’re doing phenomenally.  We have a solid roadmap for V2V and we are developing a very robust roadmap for V2I.  The Safety Pilot project which will help solidify lots of deployment questions is close to being initiated.  I believe we have solid plans that are well-articulated and transparent.  I think we are much further along than where we were 1-2 years ago. 
• You emphasized a need for systems engineering.  From presentations, pieces seem to be there, but there does not appear to be a good connection between the activities.  In a systems engineering project, there should be some contingency timing and contingency resources in case of problems or policy changes.  I don’t see this contingency present in the program plans.  Has it been considered?  Will it be built in?  Will it be done?
  • Yes, we will build it in.  We are constantly revising, and we understand your concern.  If we were to wait for perfect alignment we wouldn’t make any progress.   
• I almost see V2I as critical to V2V.  Couldn’t we install V2I at current dangerous locations and begin to offer benefits quickly?
  • We are not putting all of our eggs in one basket.  We are pushing both V2V and V2I forward.  Let’s get the technology working first.