Interpretation ID: 2669rbm

Mr. Charlie Steffens
Director, Systems Engineering
TRW Automotive
Occupant Safety Systems
4505 W. 26 Mile Road
Washington, MI 48094

Dear Mr. Steffens:

This responds to your letter requesting an interpretation of the advanced air bag requirements of Federal Motor Vehicle Safety Standard No. 208, Occupant crash protection (FMVSS No. 208). On May 12, 2000, the National Highway Traffic Safety Administration (NHTSA) published a final rule in the Federal Register (65 FR 30680) requiring advanced air bags in all passenger cars, multi-purpose vehicles, light trucks and buses with a gross vehicle weight rating (GVWR) of 8,500 lb or less. The phase-in for these new requirements begins September 1, 2003. That final rule established new, advanced air bag performance requirements to minimize the risk of injury to children, as well as new requirements to enhance protection of small and mid-size adults. The requirements in S19, S21, and S23 are designed to minimize the risk that air bags pose to infants and small children. S19 provides manufacturers with two different options for complying with the standard (low risk deployment or automatic suppression), while S21 and S23 provide three options (low risk deployment, automatic suppression, or dynamic automatic suppression). Your first question is related to the infant low risk deployment option while your three other questions are related to the interrelationship between the compliance options. I am pleased to provide a response.

You first request an interpretation of the requirement set forth in S19 and the test procedure provided in S20.4, relating to the low risk deployment option for infants. Specifically, you characterize the requirements of S20.4 as follows. "For S20.4.9, if the subject vehicle were equipped with any type of occupant sensing system that was
(1) able to recognize that the passenger seat was occupied by any infant restraint from Appendices A, B, or C of the regulation,
(2) secured into the positions and orientations described in sections S20.4.1 through S20.4.8, and a 49 CFR Part 572 subpart R 12-month-old CRABI dummy positioned according to S20.4 and
(3) the occupant sensing system and airbag control system were designed to control the output of the air bag system in a discrete manner, such as "low output", that the S20.4.9 testing would be conducted at the same discrete output determined by the occupant sensing system. Further, no additional tests to show compliance to S19, such as "high output" testing corresponding to that for a vehicle which does not have an occupant sensing system to meet S20.4.9 would be necessary."

The requirements for the infant low risk deployment option are found at S19.3, which states that "each vehicle shall meet the injury criteria specified in S19.4 of this standard when the passenger air bag is deployed in accordance with the procedures specified in S20.4." The low risk deployment option is designed to address injuries that can result when an infant is placed very close to the air bag. The risk of being directly above or adjacent to the air bag is particularly high for infants because child restraints, when placed in their rear-facing mode, will always place an infant's head close to the dashboard. A poorly installed forward-facing child seat also creates a risk, because the restraint can slide or flip forward during a crash.

S20.4 specifies several conditions for testing the deploying air bag. First, the manufacturer must assure compliance to S19.3 using any child restraint listed in subparts B and C of Appendix A to the standard. Restraints listed in subpart A (car beds) need not be tested because these restraints are not designed to be rear facing. For purposes of S19.3, the air bag is only tested with the child restraints in their rear-facing condition. This represents the worst case injury scenario. Under the specified test conditions, the vehicle seat is moved as far forward as possible, while avoiding contact with the vehicle interior. This is done to ensure that the dummy's head is placed as close to the deploying air bag as possible. The air bag is only tested with the child restraint in a belted condition.

The air bag is deployed at whatever level of force and combination of stages that would deploy in any rigid barrier crash up to 64 km/h (40 mph) when a child or test dummy is positioned in a restraint as specified in the test procedure, except that the vehicle seat may be at any seat track position. This level is determined by running an indicant test, as described in S20.4.9, at impact speeds up to 40 mph with a dummy-occupied restraint installed in the passenger seat. [1]

When NHTSA runs a compliance test on a vehicle certified to S19.3, it will only deploy the air bag at the level and, if equipped with a multi-stage inflator, with the combination of stages, that would deploy in the specified indicant test. Accordingly, vehicle manufacturers that certify to the low risk deployment option will need to ensure that their sensing systems are sufficiently robust to detect the presence of an infant in any one of the child restraints listed in subparts B or C of Appendix A.

If only a "low output" air bag deploys in the indicant test, NHTSA will not deploy a "high output" air bag simply because the "high output" air bag is placed in the vehicle for other occupants who may be seated in the passenger seat. Such an approach would have the effect of preventing vehicle manufacturers from using sensing technology to identify the presence of an infant at higher speeds.

Your next two questions relate to how the different compliance options relate to one another. Specifically, you ask if "the intent of the regulation... mean[s] that at compliance strategy needs to be chosen for each occupant size [i.e., the 1-year-old, the 3-yer-old, and the 6-year-old], but that a different strategy can be employed for each." You then ask if it is "possible to comply with ... the regulation using multiple compliance methods within an occupant size grouping using a logical subgrouping. For instance, could a system use deployment suppression for certifying the 3 and 6-year-old children in child seats and low risk deployment for the 3 and 6-year-old children not in child restraints?"

A vehicle manufacturer must certify to one of the compliance options in S19, S21, and S23. You first ask whether a different compliance option may be used for each group of children addressed by the regulation. The answer is yes. Each set of compliance options specified for a particular group of children is unique to that group. Accordingly, a manufacturer could choose to use automatic suppression to meet the infant and three year old requirements, but low risk deployment or dynamic automatic suppression for the six year old requirements. [2] However, within a given age group, a manufacturer may not choose to certify some portion of the population to one option and another portion of the population to another option. This would result in a unique compliance option that may fail to address all conditions contemplated by each option. For example, it is not acceptable to claim certification by meeting only one of the two low risk deployment positions for the three year old and 50 percent of the suppression positions. A system needs to meet one of the two options in its entirety. A customized compliance option, where the manufacturer used part of different options, but not all of any option, would create a unique compliance scheme that was never contemplated by the agency in determining how best to meet the need for safety without imposing unreasonable constraints on vehicle manufacturers. We note that in the example you provided in your letter, whereby a manufacturer would "use deployment suppression for certifying the 3 and 6-year-old children in child seats and low risk deployment for the 3 and 6-year-old children not in restraints" would be a compliant system that could be fully certified to the low risk deployment option, but not to the suppression option. Nothing in the rule prohibits manufacturers from using such a design philosophy. The deployment strategies related to children restrained in child restraints would remain within the manufacturer's discretion.

Your final question asks whether a manufacturer could use both low risk deployment and automatic suppression systems as a system redundancy. You state: "Assume that a sensing system met all of the conditions... for the RFIS low risk requirement. Suppose the sensing system also had the further capability to classify and suppress for certain tested situations. The system would be certified to meet all the low risk deployment requirements, however under certain conditions the system would suppress the airbag instead of deploying at low level because of this redundant capability. TRW believes a system of this type would comply with FMVSS 208 based upon similar NHTSA interpretations using additional or redundant safety equipment."

Your understanding is correct. Manufacturers may choose to use multiple technologies to address real world risk, without certifying to more than one compliance option. NHTSA specifically addressed this possibility in the May 2000 final rule (65 FR 30680, 30710). For example, a manufacturer may decide to use both low risk deployment and occupant sensing technologies for the six year old because it has concerns that the occupant sensing technology is insufficiently robust for all real world conditions, even though it may meet all the conditions specified in the test procedure. In such an instance, the manufacturer could certify to S23 using either the low risk deployment option or the automatic suppression option, even though in actual driving conditions, the air bag may sometimes deploy when an occupant the size of a 6-year-old is seated in the passenger seat. However, the manufacturer must choose one of the options for certification. Once it decides which option to certify to, it cannot change its position, even though the vehicle may fully meet the requirements of the other options. See S4.8. We require manufacturers to choose a particular option so that we can conduct compliance testing in an effective and productive manner.

I hope this letter addresses your concerns. Please feel free to contact Rebecca MacPherson of my staff at (202) 366-2992 should you have any additional questions.

Sincerely,

Jacqueline Glassman
Chief Counsel

ref:208
d.7/8/03