Summary
The plenary session had four invited speakers who examined thirty years of safety progress and looked at what this implied for the future. The speakers were Claes Tingvall, Director of the Swedish National Road Administration, Ulrich Seiffert, retired head of research at Volkswagen, Brian O誰eill, President of Insurance Institute for Highway Safety (IIHS) and Robert C. Lange, the Executive Director of Structures and Safety Integration at General Motors.

Professor Tingvall made a number of interesting points. Given his current position as the chairman of the EuroNCAP Board of Directors, he focused much of his talk on the role of consumer information as a means of promoting vehicle safety. He made the point that the safety of motor vehicles is improving dramatically. He stated that it was important that consumers be encouraged to purchase vehicles that perform well in EuroNCAP tests since when a vehicle is added to the fleet, it is approximately 30 to 40% safer than the vehicle that is being replaced. He further stated that a modern car is 90% safer than a 30 year old car. This is an interesting way to look at vehicle safety and why it is so encouraging that these consumer information programs have reinforced the interest in vehicle safety and has helped produce an environment where safety sells cars. And Professor Tingvall along with the other speakers believes that safety really does sell. He cited as evidence the growth of electronic stability control (ESC) in the European fleet. He also cited the growth of seat belt use reminders, in large part because of EuroNCAP ratings. The penetration of seat belt reminder systems has grown in Europe from almost none in 2002 models to an estimated 75% in 2005 models. In fact, 35% of the cars sold last year in Europe received a 5 star EuroNCAP rating. Professor Tingvall also pointed out that at least in Sweden, over 50% of new vehicle sales are to companies and public service fleets. Thus, he thinks that governments and public interest groups should work with fleet buyers to promote safety.

He voiced one concern, and that was that when vehicle manufacturers learn that EuroNCAP is considering a new rating program, manufacturers might delay development of certain technologies until they know how they will be evaluated. He was not sure if this was a major problem, but he was concerned about this issue. Another issue about which he voiced concern was how to evaluate new crash avoidance technology, particularly for consumer information programs. He made the point that until technology is introduced into the fleet, we really have no way of determining just how effective it will be. And because we do not conduct multi-center international evaluations, it takes far longer than it should to evaluate new technology as it moves in the fleet. He cited ESC as an example where it took five years to develop effectiveness estimates for the technology because researchers did not work together and pool their data to get sample sizes that would have been large enough to produce effectiveness estimates much sooner.

And for crash avoidance technology that is not yet on the road, we do not have the tools to develop reliable predictions of effectiveness that could serve as the basis for rating programs or for rulemaking. Professor Tingvall thought that one possible way to address this might be for manufacturers to develop test procedures that could be used to evaluate crash avoidance technology and then convince governments or NCAP Boards of the validity of their approach. However, Brian O誰eill commented that if this scenario was in place when antilock brakes were first being introduced, the safety community would have probably agreed that ABS would significantly reduce crash frequency, something that never materialized.

Professor Seiffert also agreed that safety sells, buttressed by pressure from governments, insurance companies, and NCAP programs. But he thought that the most effective way to improve safety was through natural competition between safety engineers at the suppliers, the manufacturers and at research institutes as they respond to this pressure. He believed that promoting this competition between innovative engineers was a far more effective way to produce advances in safety than was regulation. And like Professor Tingvall, he thought that the premature development of standardized test procedures could stifle technology development. He specifically voiced concern that if a simplified test procedure was used to mandate ESC, in the end it could retard the development of more advanced systems. And coming from a career in vehicle manufacturing, while he agreed that safety was important, he noted that it was only one of many factors that must be considered in vehicle development. He was also the only panelist who noted that products liability exposure was a serious impediment to the introduction of advanced technology, particularly advanced crash avoidance technology.

Professor Seiffert identified a number of areas where he expected to see safety improvements. They included adaptive restraints, compatibility, including occupant compartment integrity, pedestrian protection, rollover injury mitigation coming from inflatable curtains, fuel system integrity, collision notification systems, automatic braking, although its introduction will be delayed by products liability concerns, pre-safe systems and better accident data. And he thought that in many cases regulations inhibit the adoption of advanced technology, and moreover, once regulations are in place, they never go away, regardless of whether or not there is a continuing need for them.

Brian O誰eill also believes that vehicle rating programs lead to better vehicle designs. He cited clear evidence of crashworthiness improvements in response to the IIHS front and side impact evaluations. And he believes that in many cases, the industry response to consumer information programs is quicker than their response to regulation. But he is not optimistic about our ability to produce similar results with crash avoidance ratings. He raised serious questions about our ability to evaluate active safety technology before it is in the fleet. And even if we can determine that some specific countermeasures might be effective, he foresees major challenges in being able to differentiate between versions of crash avoidance technology to produce meaningful ratings.

Further, Mr. O誰eill felt that there is very little good research being done to develop the scientific foundation that can be used to evaluate crash avoidance technology. And he thinks that our data systems do a poor job of identifying crash avoidance issues.

Bob Lange was the final panelist and he opened his presentation with a description of the industry痴 financial environment. He contends that given today痴 overcapacity, more than the minimum level of safety is a requisite for entry into the market. And because of pricing pressure on manufacturers, they cannot get price increases for this enhanced safety content. He believes that these pressures, coupled with consumer rating programs, will continue to produce advances in vehicle safety. But because of these financial pressures, he believes that it is imperative that we develop better science to allow the evaluation of crash avoidance technology and better data to focus our vision. He argues that this is the only way to provide the resources for continued improvements in vehicle safety.

He cited the GM study in which they identified 44 crash types that accounted for virtually all crashes. An analysis using these findings was used to predict that stability control systems were viable countermeasures for many of the crash types identified. But while this provides directional information, it does not provide real precision. And that is why Lange argued that we need better data and better science.

With respect to future technology, he thinks that we can expect to see frontal collision warning systems before too long and he predicts that they will reduce the frequency of rear end collisions.

Conclusion
Overall, the speakers were unanimous that consumer information programs are advancing safety, and that the market now rewards safety improvements. And while they all agree that there is a great deal of advanced crash avoidance technology in the future, they were all concerned that our ability to evaluate the potential benefits of these devices was at best, directional. Our data systems are not yet capable of doing an adequate job of characterizing the causes of most crashes, and our scientific tools are not adequate to predict just how drivers will respond to advanced safety systems, and hence, we are not in a position to accurately predict safety benefits. Further, we are not even sufficiently organized to conduct collaborative investigations of new technology when it does move into the fleet, so reliable estimates of technology effectiveness take far longer to produce than they should.

But overall, during the last thirty years the panel agreed that we have seen significant advances in safety and there is no reason not to expect this progress to continue into the future.