Ms. Chaka presented document GTR9-6-17, explaining the issues that are seen for the compliance of heavier trucks and SUV with the pedestrian safety requirements. Ms. Chaka explained that several customer requirements and needs prevent the vehicles from meeting the legform requirements or that the vehicles would no longer be able for several of their original purposes. In addition, many of these vehicles would need to comply with both, the FlexPLI as well as the upper legform to bumper test requirements due to their specific design. This seems to be impossible from an engineering point of view.

Mr. Kinsky reminded the group that already during the discussion defining the gtr No 9 requirements it had been highlighted that the pedestrian safety tests that are used today have just been developed for the typical (sedan-type) vehicles and similar vehicle-designs.

The chair noted that for other vehicle categories no further information was provided. Obviously, no issues with the design of such vehicles towards FlexPLI requirement are seen. He therefore proposed to close the respective action item A-3-12 which was agreed. Also, the chair noted that the wording of the preamble will consequently reflect the issues mentioned for specific vehicles.

The Alliance had already presented in earlier meetings to see issues with the cost benefit assessment provided to the Informal Group. Mr. Bilkhu now specifically addressed the concerns of the US manufacturers with the two documents GTR9-5-14 and GTR9-5-19. These concerns were reported in a study of JP Research, a contractor of the Alliance (document GTR9-6-16). Mr. Bilkhu presented the summary (document GTR9-6-15r1) of JP Research’s findings. He concluded that the methods used by JASIC and BASt may potentially lead to unrealistic assessments of the benefit for the US. On request of Mr. Martin he confirmed that “unrealistic assessment” in this case means “overestimation”.

Mr. Martin wondered whether the Alliance can draw different conclusions from their project. After some discussion on this it was noted that further work on this may be needed at least for future US rulemaking as NHTSA does not fully agree with the all assumptions made by the Alliance.

Mr. Zander commented that estimation of a benefit is a common process and may not be correct in all cases. However, benefits can just be estimated and a certain approach has to be used for this. BASt made good experiences with shifting the injuries and this is a well-known process that is widely accepted. Also, Mr. Zander pointed out that the assessment of JP Research may not have used the right figures of the GIDAS database. He presented a short document on this (document GTR9-6-25). Mr. Bilkhu promised to double-check the information. Mr. Edwards wondered whether it is an appropriate approach to shift injuries by one AIS and Mr. Zander confirmed that this is based on estimations. However, it needs to be noted that only those injuries are shifted where an improvement can be expected by the new tool. This does not necessarily lead to a shifted MAIS. After some discussion it was finally agreed that the Alliance will double-check the comments of BASt and provide feedback before the next meeting; all other parties are invited to also do so (action item A-6-05).

Dr. Konosu added that the data summarized by JASIC are Japanese data and that they are seen to be acceptable for Japan. However, it is true that the US may have a different situation. Then, the US should assess their data but cannot simply compare Japanese data to the PCDS data. Mr. Bilkhu agreed that an analysis for the US is needed and explained that this is usually delivered by their authority. Mr. Edwards added that the intention of the Alliance is just to get a better understanding of the assumptions made by JASIC.

On request of the chair, Mr. Stammen explained that NHTSA for the time being has no further comments on the cost benefit analysis. He explained that the preamble can reflect the discussion above which would be sufficient for NHTSA. Mr. Martin added that all presentations made were very valuable for NHTSA and will be considered in their rulemaking process. It was therefore concluded that action item A-5-09 from the 5th meeting can be closed.

Mr. Bilkhu added that, however, the US may have certain specifics that anyway may lead to different assessments of the cost benefit issue. Specifically, the bumper requirements in part 581 are in contradiction to the pedestrian safety requirements and – responding to a question of Mr. Broertjes on this in between – at the moment no vehicle complies to both, part 581 and (European) pedestrian legform test requirements. Some discussion on this came up, also mentioning that similar issues could exist with future UNECE-R 94 (frontal impact) requirements that currently are in discussion in Geneva. It was finally concluded that rulemaking processes in the US need to consider all these details in their future activities.

To finalize the discussion on this subject for the time being, Dr. Konosu presented a summary of his comments (document GTR9-6-27). He stated that the intention of Japan was to prepare information for their country and that other attendees of this meeting are very welcome to do the same for their countries.

The Alliance had already presented in earlier meetings to see issues with the cost benefit assessment provided to the Informal Group. Mr. Bilkhu now specifically addressed the concerns of the US manufacturers with the two documents GTR9-5-14 and GTR9-5-19. These concerns were reported in a study of JP Research, a contractor of the Alliance (document GTR9-6-16). Mr. Bilkhu presented the summary (document GTR9-6-15r1) of JP Research’s findings. He concluded that the methods used by JASIC and BASt may potentially lead to unrealistic assessments of the benefit for the US. On request of Mr. Martin he confirmed that “unrealistic assessment” in this case means “overestimation”.

Mr. Martin wondered whether the Alliance can draw different conclusions from their project. After some discussion on this it was noted that further work on this may be needed at least for future US rulemaking as NHTSA does not fully agree with the all assumptions made by the Alliance.

Mr. Zander commented that estimation of a benefit is a common process and may not be correct in all cases. However, benefits can just be estimated and a certain approach has to be used for this. BASt made good experiences with shifting the injuries and this is a well-known process that is widely accepted. Also, Mr. Zander pointed out that the assessment of JP Research may not have used the right figures of the GIDAS database. He presented a short document on this (document GTR9-6-25). Mr. Bilkhu promised to double-check the information. Mr. Edwards wondered whether it is an appropriate approach to shift injuries by one AIS and Mr. Zander confirmed that this is based on estimations. However, it needs to be noted that only those injuries are shifted where an improvement can be expected by the new tool. This does not necessarily lead to a shifted MAIS. After some discussion it was finally agreed that the Alliance will double-check the comments of BASt and provide feedback before the next meeting; all other parties are invited to also do so (action item A-6-05).

Dr. Konosu added that the data summarized by JASIC are Japanese data and that they are seen to be acceptable for Japan. However, it is true that the US may have a different situation. Then, the US should assess their data but cannot simply compare Japanese data to the PCDS data. Mr. Bilkhu agreed that an analysis for the US is needed and explained that this is usually delivered by their authority. Mr. Edwards added that the intention of the Alliance is just to get a better understanding of the assumptions made by JASIC.

On request of the chair, Mr. Stammen explained that NHTSA for the time being has no further comments on the cost benefit analysis. He explained that the preamble can reflect the discussion above which would be sufficient for NHTSA. Mr. Martin added that all presentations made were very valuable for NHTSA and will be considered in their rulemaking process. It was therefore concluded that action item A-5-09 from the 5th meeting can be closed.

Mr. Bilkhu added that, however, the US may have certain specifics that anyway may lead to different assessments of the cost benefit issue. Specifically, the bumper requirements in part 581 are in contradiction to the pedestrian safety requirements and – responding to a question of Mr. Broertjes on this in between – at the moment no vehicle complies to both, part 581 and (European) pedestrian legform test requirements. Some discussion on this came up, also mentioning that similar issues could exist with future UNECE-R 94 (frontal impact) requirements that currently are in discussion in Geneva. It was finally concluded that rulemaking processes in the US need to consider all these details in their future activities.

To finalize the discussion on this subject for the time being, Dr. Konosu presented a summary of his comments (document GTR9-6-27). He stated that the intention of Japan was to prepare information for their country and that other attendees of this meeting are very welcome to do the same for their countries.

Report evaluating the methodology and use of available accident data in documents GTR9-02-07r1 and GTR9-02-12 and providing recommendations on how the available field data could be used to estimate the number of fatalities/injuries associated with tibia fractures.

The chair explained that the cost-benefit assessment appears to be a major topic in this group. JASIC elaborated a study on cost benefit (documents GTR9-2-07r1 and GTR9-2-12). During the 2nd and 3rd meeting the Alliance expressed orally their concerns related to the accident database used in the JASIC papers. For the 4th meeting the Alliance submitted two documents (documents GTR9-4-12 – report and GTR9-4-13 – presentation) as technical information to the group as result of a request to JP Research as their contractor to assess, among others, the appropriateness of several accident data sources (PCDS – Pedestrian Crash Data Study, FARS – Fatality Analysis Reporting Systems, NASS/CDS – National Automotive Sampling System/Crashworthiness Data System etc., including their combinations) and the use MAIS levels 4+ for estimating leg fractures and fatalities. JP Research found some differences in the case numbers used by JASIC. In addition, JP Research stated that the PCDS data set is not representing the current US fleet and appears to be statistically not valid. There are also remaining concerns regarding the combination of the existing data.

JASIC clarified the different case numbers in their study. JASIC omitted those cases with fractures in the upper tibia plateau and in the ankle area and explained that the focus was on tibia shaft fractures only.

In addition BASt presented a benefit estimation of FlexPLI vs. EEVC LFI (document GTR9-4-18). BASt concluded that the injury assessment ability using the FlexPLI is better than using the EEVC LFI due to the direct measurements of ligament elongations and bone bending. For example, the fracture risk can be measured at 4 locations along the tibia bone shaft for the FlexPLI while the fracture risk judgment of the EEVC LFI concentrates to the upper tibia part. BASt stated that cars which fulfill the FlexPLI requirements should also pass the tests with the EEVC LFI. OICA opposed that this is not necessarily true for all vehicle types. Japan stated that the better biofidelity of the test tool should be decisive for those questions.

JASIC presented the Honda document for the 2012 IRCOBI conference (held in Dublin from 12th – 14th September 2012) which was agreed to be circulated as document GTR9-4-20 to the group. JASIC concluded among others that the correlation with human injury measures was found to be significantly improved for FlexPLI relative to EEVC legform for tibia fracture and ACL failure measures.

JASIC stated that Japan had made its study on cost-benefit and that no further work is planned for this issue. The chair proposed to have a deeper look into the GIDAS data and requested BASt to provide some additional information (action item A-4-11) by the next meeting. Since this issue is also of high priority for NHTSA (see agenda item 5) the chair additionally proposed to contact with NHTSA to find a common solution (action item A-4-10) for the next informal group meeting. NHTSA will undertake additional studies.

Alliance of Automobile Manufacturers presentation on the results of the JP Research inquiry into the propriety of using PCDS and NASS/CDS data to derive injury estimates by AIS levels for pedestrians.

The chair explained that the cost-benefit assessment appears to be a major topic in this group. JASIC elaborated a study on cost benefit (documents GTR9-2-07r1 and GTR9-2-12). During the 2nd and 3rd meeting the Alliance expressed orally their concerns related to the accident database used in the JASIC papers. For the 4th meeting the Alliance submitted two documents (documents GTR9-4-12 – report and GTR9-4-13 – presentation) as technical information to the group as result of a request to JP Research as their contractor to assess, among others, the appropriateness of several accident data sources (PCDS – Pedestrian Crash Data Study, FARS – Fatality Analysis Reporting Systems, NASS/CDS – National Automotive Sampling System/Crashworthiness Data System etc., including their combinations) and the use MAIS levels 4+ for estimating leg fractures and fatalities. JP Research found some differences in the case numbers used by JASIC. In addition, JP Research stated that the PCDS data set is not representing the current US fleet and appears to be statistically not valid. There are also remaining concerns regarding the combination of the existing data.

JASIC clarified the different case numbers in their study. JASIC omitted those cases with fractures in the upper tibia plateau and in the ankle area and explained that the focus was on tibia shaft fractures only.

In addition BASt presented a benefit estimation of FlexPLI vs. EEVC LFI (document GTR9-4-18). BASt concluded that the injury assessment ability using the FlexPLI is better than using the EEVC LFI due to the direct measurements of ligament elongations and bone bending. For example, the fracture risk can be measured at 4 locations along the tibia bone shaft for the FlexPLI while the fracture risk judgment of the EEVC LFI concentrates to the upper tibia part. BASt stated that cars which fulfill the FlexPLI requirements should also pass the tests with the EEVC LFI. OICA opposed that this is not necessarily true for all vehicle types. Japan stated that the better biofidelity of the test tool should be decisive for those questions.

JASIC presented the Honda document for the 2012 IRCOBI conference (held in Dublin from 12th – 14th September 2012) which was agreed to be circulated as document GTR9-4-20 to the group. JASIC concluded among others that the correlation with human injury measures was found to be significantly improved for FlexPLI relative to EEVC legform for tibia fracture and ACL failure measures.

JASIC stated that Japan had made its study on cost-benefit and that no further work is planned for this issue. The chair proposed to have a deeper look into the GIDAS data and requested BASt to provide some additional information (action item A-4-11) by the next meeting. Since this issue is also of high priority for NHTSA (see agenda item 5) the chair additionally proposed to contact with NHTSA to find a common solution (action item A-4-10) for the next informal group meeting. NHTSA will undertake additional studies.