The Chair questioned how the manufacturers could do today between ITS, LDWS, ISO, Vienna Convention, product liability, UN R79, for approving a vehicle.

Japan presented the document LKAS-01-04. The expert from Japan informed the group that this document was an attempt to cover Industry’s concerns with regard to LKAS.

OICA found that a column showing the items covered by UN R79 was missing (warning, failure modes, whether in the core of the text or in the CEL annex).

The expert from OICA pointed out that each item was covered in some way or another by UN R79, ISO or the current Japanese guidelines. The delegate was seeking clarification about the “purpose” column in the document LKAS-01-04, e.g. the purpose of provisions for operating speed. Japan found this similar to the provisions in the introduction of AEBS regulation. Japan was keen that some fruitful discussions take place on these items.

Japan presented document LKAS-01-05. The expert from J clarified that in general, the option numbering is from the most severe (1) to the most relax (4).

The Chair found it necessary to study in depth the documents because it shows that, even if the LKAS is covered by different texts, their interpretation is such that the level of performance can differ very much (even be opposite).

Japan was also keen to discuss the document in depth, then decide the way to proceed.

It was suggested to add a column in document LKAS-01-05 showing how UN R79 addresses each item. Conclusion: the group agreed to add a column, for assessing whether the items are covered by R79. The aim was that, at the end of the exercise, the group should decide the way to proceed, i.e. amending UN R79, elaborate new guidelines, or any other option. The Secretariat then created the document LKAS-01-05-Rev.1

Definitions
A debate took place about the definitions in UN R79 with regard to the different systems. Paragraph 2.3.4. specifies that “the driver remains at all times in primary control of the vehicle” i.e. the driver must control the steering control (in the frame of the steering system regulation); the secondary control of the vehicle is provided by systems like LKAS. J found the definitions sufficient as

LKAS is clearly a corrective steering system. Yet at the time of autonomous steering system, it may be opportune to improve the definitions of UN R79. The European Commission contested the reading of UN R79, because of the texts of paragraphs 2.3.4.1. and 2.3.4.2., where PAS is part of advanced driver assistance, without the driver having the hands on the steering wheel.

Scope
Japan was keen not to exclude any category, but proposed to focus on M1 w/o excluding any other category. In addition, the expert from Japan informed that LKAS currently requires electric steering system.

OICA clarified that heavy vehicles are currently implementing LDWS, and do not have LKAS. Still OICA found LKAS not mature for a rulemaking because on M1 vehicles, different systems are still under development. Yet the concern of a wide scope would be that a Contracting Party mandates LKAS on vehicles fitted with hydraulic steering.

CLEPA was confused that the figures of the Japan documents justified LKAS on heavy vehicles, while Japan proposes that the ad-hoc group focuses on LKAS for M1 category.

Japan clarified that data only exist for heavy vehicles, and they extrapolated them for light vehicles.

Operating speed
Japan suggested option 2 because the key parameter is to maintain road safety (avoid driver’s confusion).

Japan announced being flexible about minimum performance requirements (e.g. operating speed limit), i.e. would agree deleting.

Japan justified the 60km/h limit of option 2 per the limit in LDWS regulation, and LKAS is designed for primarily use on highway. OICA found it necessary that flexibility be provided in operating speed limit, because there is no clear limit in
state of the art systems. CLEPA challenged the proposed wording (shall) because this may prevent the driver from the control of the vehicle. J conceded that this was subject to the presence of a manual switch. OICA clarified that it is
impossible today to determine an operating speed limit and that there is a need for a manual switch because this is no emergency system.

Conclusion: item not covered by UN R79. Group to decide later about possible speed limit

Acceleration etc. caused by the operation of the system
Japan commented that this item is necessary to avoid driver’s confusion, but Japan was ready to hear the other delegations’ opinions.

OICA pointed out that the item raised in document LKAS-01-04 is covered by UN R79, yet the proposal from J per document LKAS-01-05 is not covered the regulation.

The Netherlands found that the lateral acceleration is covered by UN R79, but the longitudinal acceleration is covered by UN R13. In addition, a good understanding of the CEL annex would prevent dangerous systems, even while no value is specified.

The group was informed that UN R13H by selective braking permits deceleration up to zero.

CLEPA was of the opinion that the requirement that the driver can always override the system would cover this item because the driver can simply apply the necessary force on the steering wheel.

Conclusion: need for further discussions on this item.

Road shape
OICA found the value of 1000m reasonable.

The Chair pointed out that having a small curve limit would mandate the system to detect lane markings up to this limit. OICA stressed that this was covered by one of the comments raised by OICA at GRRF, that a limit would prevent systems designed for straight road.

CLEPA added that a too low limit would not be convenient to the driver on country roads.

Japan supported a clear distinction between LDWS and LKAS as LKAS in intrusive. Japan also supported the 1000m limit.

The Netherlands found that for a safety systems, it does not matter whether it is comfortable, because if the driver is not happy, then he can switch off the system.

Lane marking
Japan clarified their proposal for option 2 because of the need for early detection by LKAS; the expert indicated the need for time and space for initiating control of the steering system. He said that, in addition, state of the art systems are different according to the markets. J agreed that the wording should be improved.

Industry indeed challlenged the proposed wording.

Korea informed that the country has blue lane markings.

The Netherlands recalled that the UN regulations must cope with mutual recognition and that it would be counterproductive to have provisions limiting this concept.

OICA pointed out the technical challenge behind the discussion Germany repeated the need not to exclude systems which can function without lane marking. Japan was of the opinion that such systems would be out of scope. CLEPA pointed out that
the absence of markings on certain portions would make the system switch from activation to deactivation repeatadly with inconveninece for the driver.

Japan stressed that the current situation is that a system approved per UN R79, could face approval problems in Japan.

Conclusion: need for further discussions.

LKAS performance requirement
Should some criteria be adopted, some amendments to UN R79 or other regulation should be initiated.

Conclusion: not covered by UN R79.

Functional limitation of the system
Japan found this item necessary for avoiding confusion. The group agreed to pay attention to the meaning of “operating”.

CLEPA challenged the reference to “high priority warning” because it is not yet sure whether LKAS discussed in this group addresses safety system or comfort system.

Conclusion: need for further discussions.

Requirement for the end of the system operation
OICA recalled that the CEL annex covers this for avoiding roll-over, etc. “easy and safe handling of the vehicle up to maximum speed”.

The Chair raised the concern of the situation when the system looses the marking within a curve: how does it give back the control to the driver?

CLEPA clarified that the end can be done smoothly in the case the marking cannot be detected anymore. Yet if the system becomes out of order (failure), then the end should occur abruptly.

The Netherlands said that a Technical Service should check per CEL annex, assessing what would be the consequences if such event would occur, and added that a Technical Service may request to perform a test. The criterion would be that the
vehicle remains stable, yet UN R79 does not contain any test nor value for this.

Germany informed that such systems abruptly releasing the control do exist, yet the forces/moments are so low that they do not make the vehicle unstable.

CLEPA confirmed that the systems currently in the market do not have safety issues thanks to the level of torques.

OICA also recalled that UN R79 is clear that the automatic systems, per Annex CEL, cannot cause any deterioration of the steering system handling.

The group convened that there is no intention to bring suspicion on the Technical Services, and was keen to provide assistance when felt necessary.

Holding a steering wheel by the driver
The Netherlands was keen that the system avoids overreliance. The expert favoured option 1 (the operation shall be cancelled when there is no steering operation of the driver for more than 5s) + an optical warning. The expert also questioned the 5 second value.

CLEPA was of the opinion that a sleeping driver cannot remove the hands from the steering wheel.

OICA questioned whether switching off the system when the driver is asleep would always be a good point for safety.

The Chair supported that point of view because the system may give the conditions for sleeping, then switch off the assistance.

Japan favoured option 3 because the Vienna Convention is clear that the driver should not be asleep. The expert from J informed that J assumes that the driver controls the vehicle even if the hands are not on the steering wheel.

CLEPA suggested that the warning be repeated if there is no reaction from the driver.

OICA pointed out that there must be a balance between the avoiding over-reliance and too frequent alarms.

Conclusion: not covered by R79.

NL, supported by the European Commission, found the criterion important because the assistances should not provide the driver too much tools to leave the control to the system.

Korea found this important for safety. The expert favoured haptic or acoustic warning.

Germany found a warning necessary, with no suggestion

Spain had no opinion.

Conclusion: important item, partially covered by UN R79.

Override
Conclusion: Fully covered by UN R79.

Condition of non-operational being allowed
There was a debate about whether actionning the direction indicator can be considered as “overriding” the system.

Conclusion: Option 3 favoured by the group and paragraph 5.1.6. of UN R79 addresses this.

ON/OFF switch
The group was informed about the discussions currently being held at WP1 that a system not covered by a UN regulation must be overridable or capable of being switched off.

CLEPA would recommend default off, with a mandatory switch that the driver can voluntarily switch on the system.

The European Commission would favour a default on, with a switch such that the driver can switch off the system.

Some CLEPA experts pointed out that LKAS is a system internening when normally driving, hence would recommend default off.

Japan justified their position that almost all current LKAS are “last memory” systems and Japan favoured a default off system.

OICA stressed that the criterion is the acceptability from the driver, a default on system is acceptable if it does not burden a driver. The expert suggested not to be too restrictive today, until the time there are lots of systems in the market.

The group was informed that in addition, the “switch” can be a software button in a menu.

The Netherlands thought that a comfort system should be default off with a switch.

Korea found that a mandatory system should be default on.

The European Commission did not find important whether it is default on or off, and supported a switch and would favour default on system. Germany fully supported the European Commission.

Conclusion: need for further discussion

Malfunction warning / Status display
OICA clarified that UN R79 does cover this per the CEL annex.

Japan clarified that “operating” means passive activation and was keen to distiguish functional from internal failures.

CLEPA did not find beneficial for the driver that the system indicates whether it can detect the lane markings, rather whether it can intervene.

The European Commission found sufficient that the system indicates whether there is a malfunction (electric, sensor, etc). The rest was considered not relevant as mainly for maintenance purpose.

Conclusion: covered by R79, + need for more discussions.

Information to be known to users
Japan was not keen for such requirement.

OICA pointed out that the document reflects the current situation. UN R79, paragraph 5.5.2. It must be possible to verify in a simple way the correct operational status of those Electronic Systems, which have control over steering. If special information is needed, this shall be made freely available. This is only for PTI.

The Secretariat was of the opinion that the driving license covers this.

Conclusion: no need for requirement.

Failsafe

CLEPA supported that there is a request for redundancy.

Conformity with the safety aspects of complex electronic control systems

Japan supported the current wording of UN R79

There was no comment from the group.

EMC
Acceptable, no comment.