Avvakta med att köpa FSD. Det är troligen pengarna helt i sjön på en bil som man bara ska behålla i tre år.
https://wiki.unece.org/display/trans/AD ... th+session5. Drafting the new UN Regulation on DCAS
The Chair invited the Secretary to provide a review of the Master Document (ADAS-18-02). The Chair thanked the Secretary for the review and requested the Secretary to create a list of items that would require further review by other UN working parties or groups. (e.g. with regard to requirements on the driver, e.g. confirmation of knowledge).
OICA-CLEPA requested that any new textual proposals not discussed by the SDG would be shared ahead of time with notification.
The Secretary confirmed that the leadership will aim to do so in the future.
ITU indicated that any requirements for interfacing between WP1 and WP29 should be discussed during the WP.29 and AC.2 session taking place in two weeks, and urged the leadership to inform the GRVA Chair.
The Chair noted that this might be too early as more detailed comments would need to be provided. TF ADAS needs more time to clarify to outlying issues; however, the Chair agreed with ITU to inform the GRVA Chair ahead of the upcoming discussion at WP.29 and AC.2 in March.
The Chair invited France to present ADAS-18-04 regarding the two-step approach to regulating DCAS systems.
France explained that currently there would be two means for approving level 2 systems, either through Regulation No. 79 or the DCAS regulation. As a first step, France would propose to differentiate both regulations by way of the means of activation. For Regulation No.79 or No.13 H, systems would be addressed that have a separate means of activation, whereas DCAS would incorporate systems with a unique means of activation. France’s proposal aims to resolve potential confusion in terms of type-approval and implementation of the systems. In the second step, all L1 and L2 systems would be regulated under DCAS.
The Chair thanked France for the proposal which raises an important question in terms of interaction between DCAS and other regulations, and indicated the relevance of this proposal to the question raised by the GRVA Chair about the further development of UN Regulation No.79. The Chair invited any textual proposals to fit the DCAS regulation according to the first stage, and subsequently as the second stage.
Japan thanked France for the important proposal. Japan would support the two-step approach especially with regard to the second step as the final goal where level 2 systems would only be approvable under a single regulation. The most important thing is the first step, which Japan is concerned about. Japan suggested as an alternative proposal to maintain the current Regulation No.79 and to have DCAS only address L2 systems not addressed in the first step.
France clarified that means of activation is intended only for L2 in DCAS in the first step, whereas Regulation No.79 would require individual activation each time.
OICA-CLEPA thanked France and Japan for the very clear discussion. OICA-CLEPA expressed that we should not solve all issues at once and that separation by means of activation may not be appropriate, as the system configuration could allow a system to fall in both categories. OICA-CLEPA indicated that by building the requirements for systems not regulated by Regulation No.79, sufficient understanding as building blocks might be identified which should potentially allow amendment of other regulations accordingly.
France thanked OICA-CLEPA but noted that the generic requirements in DCAS may risk unharmonized type-approval of systems when compared to other regulations.
ITU noted that stakeholders should consider that the ODDs may be different even though there might be functional similarities between systems. As an example, a driver may choose to use solely lane centering or to use DCAS on a highway.
The Secretary reminded that the Terms of Reference required in any case an evaluation of interaction between DCAS and other regulations at a later stage. This parallel activity could be started in a few months depending on the maturity of the master document.
The Chair thanked all for the input.
The Chair invited ETSC to introduce ADAS-18-03.
ETSC noted that the comments were provided on document ADAS-16-04 rather than the latest version. ETSC is concerned about the direction that the DCAS Master Document is heading in. ETSC had hoped for a more robust document to be developed as a follow-up to Regulation No.79, taking into account considerations of the Dutch Safety Board report. Regulation No.79 also has severe shortcomings which ETSC should not be considered as a basis. ETSC expressed strongly that further research should be considered as the safety benefits of these additional ADAS capabilities has not been evidenced. Adding more and more automation to the driving task leads to the driver increasingly becoming an operator. Driver monitoring systems may not be reliable enough to consistently detect driver distraction as indicated by industry during EU-level drafting discussions on driver monitoring. ETSC called for regulatory activity on driver monitoring systems.
The Chair noted that DCAS should only be enabled through effective driver monitoring to ensure engagement in the driving task. The manufacturer would have to ascertain reliable performance of this system in order to enable type-approval and/or deployment.
EC expressed interest to listen to the IIHS studies. EC noted that not just driver distraction is relevant, but also the ability of the driver to respond or correct is important. In this sense, reaction times (e.g. in the context of hands-off time) will need to be considered. This will need to be addressed in the document.
The Chair noted that the ability for the driver to respond needs to be addressed and described by the manufacturer as part of the system’s control strategy documentation.
The EC noted that this was already addressed in part in the current text in a general sense and may need further review, e.g. regarding failure (not just electrical) of systems.
OICA-CLEPA noted that ETSC comments consistently require more detailed requirements or clarification of higher level terms, but expressed that this was exactly what was intended with the development of the DCAS regulation avoiding the limitations of Regulation No.79. The balance that is being developed through the Audit process of detailed assessment of operational and functional safety requiring appropriate processes are put in place in the context of system design. OICA-CLEPA indicated this is crucial as this would otherwise require consistent iteration of the regulation to allow for specific system design. Regarding the comment that some system systems cannot be approved, OICA-CLEPA indicated to have been very clear from the start about intent. OICA-CLEPA suggested to refocus the discussion away from barring specific systems from type-approval, to defining appropriate requirements that would ensure safety and would allow such systems to be approved. Lastly, regarding the establishment of performance requirements due to imperfect performance of a given system, OICA-CLEPA reminded that these systems will require the continued engagement of drivers which is an important separator with ADS. OICA-CLEPA would be happy to address specific concerns that can be resolved by developing additional requirements.
The Chair reminded the TF that the multi-pillar approach for assessment will be used which include audit and monitoring to help maintain assurance of the safety of the system.
The UK commented that the use of subjective terms is difficult in the context of a regulation that requires mutual recognition as this may be a subjective assessment. A consistent understanding of those terms is desirable in this context, as driving behaviors may differ between different markets. The challenge with DCAS may be harder as it may not be understood how the driver would respond. The UK would welcome information on how these systems are being used and managed by the driver. Regarding driver engagement and sufficiency of requirements, it may not yet be conclusive.
The EC thanked OICA-CLEPA and the UK for the comments, and shared many of the concerns while recognizing the need of industry. The EC noted that the regulation should be selective and repeatable enough to differentiate good from bad designs, and ensure consistent enforcement. The EC asked the industry to consider how the regulation could be made sensitive enough to ensure reproducibility. Some features or characteristics may need to be guaranteed, where an ADS approach cannot be maintained.
OICA-CLEPA noted that the problem with parameters and thresholds is that they often apply to a limited scenario, and once you leave that, they no longer work. So either we find another way of ensuring safety than establishing fixed regulatory thresholds, or we should at least create the flexibility in the thresholds we define to cover all relevant driving situations.
The EC commented that the system should be robust as well.
The Chair noted that repeatability and reproducibility are not easy to resolve. The Chair expressed the hope that the Audit of defined criteria would be sufficient to ensure a shared understanding between authorities. If not, specific workshops similar to those organized in the context of Regulation No.155 may need to be organized.
ETSC thanked feedback so far and expressed that ETSC would begin contributing to the SDG. Regarding the comment that some systems are not liked, ETSC mentioned that this position is based on available scientific evidence or comments received from consumers, including from markets where such systems are deployed. Research evidence, including from other fields, indicated that consumers are not good monitors of systems.
The Chair welcomed ETSC’s representative to the SDG and commented that a list of systems subject to regulation may be desirable.
OICA-CLEPA mentioned that this in part is already addressed through the specific features listed in the document, and that work may be focused on a subset of these features. OICA-CLEPA also noted that a list of use-cases was previously provided (ADAS-08-04).
The Chair invited OICA-CLEPA to introduce ADAS-18-05.
OICA-CLEPA explained that there are some sections where textual proposals still need to be provided. Regarding 5.1.5., OICA-CLEPA explained that they struggle to establish a link with other regulations considering such systems are mandated at the national level.
The EC explained that the IIHS study referenced that ADAS is ideally coupled with emergency safety systems. For this reason, EC expressed preference to require the presence of emergency safety systems as a conditional to enabling the use of DCAS, for example through reference to other regulations in the master document.
OICA-CLEPA inquired whether it may not be sufficient to require that emergency safety systems fitted on the vehicle cannot be manually deactivated in order to enable DCAS use. OICA-CLEPA expressed that this would avoid the challenge raised due to direct reference to other regulations.
The EC indicated willingness to consider but invited input from other stakeholders.
The UK commented that more complex systems may require a more extensive set or capability of safety systems to be installed.
Japan supported the EC’s statement and expressed preference to have emergency safety systems installed as a conditional for DCAS.
The Chair thanked the comment and referred the matter to the SDG for further discussion.
On 5.4.1.2, OICA-CLEPA questioned whether there is a need to maintain the provision considering the lane keeping requirements provisions further in the document. In addition, OICA-CLEPA suggested to remove reference to ‘relevant for the system’s intended capability’.
The EC accepted the proposals for 5.4.1.1 and 5.4.1.2 as it would enable coverage of more systems.
Regarding 5.4.1.8., OICA-CLEPA suggested improvements to address the fact that the system is level 2 rather than automated.
The EC felt the changes were acceptable, but requested OICA-CLEPA to clarify what is defined as a failure. Regarding controllability, the term was accepted however the EC requested to clarify specifically that this includes consideration for driver reaction time and ability. In terms of hands-off, the EC suggested that hands-off systems are further considered at a later time.
The UK agreed with the EC that failures need to be further specified in more detail as the proposal provides less coverage than what was intended with initial text. The UK also agreed with the EC in terms of controllability.
OICA-CLEPA clarified that this was not intended to be excluded through the wording. OICA-CLEPA confirmed that they could come up with a term as needed, but urged to stay away from explicit detail on failures.
The EC offered to draft text, however suggested to resolve the matter by simply adding reference to ‘malfunction’ in the master document text.
The Secretary additionally suggested to address the meaning of ‘failure’ or ‘malfunction’ in the definition section.
On 5.4.3.6, OICA-CLEPA explained that they did not see a need to regulate longitudinal control considering the good performance up to this point. OICA-CLEPA suggested to focus specifically on any deceleration or longitudinal control requirements that are necessary for individual maneuvers or certain features.
The Chair proposed to follow OICA-CLEPA’s suggestion to move consideration of longitudinal control to the section on specific features.
The EC could not confirm a position at this time, and noted that there are aspects of longitudinal control (e.g. following distance) which may not be relevant to specific types of DCAS features. In particular, the EC felt that consideration is required of configurability of DCAS behavior and longitudinal control by the driver.
The Chair invited textual proposals to address and to remove placeholders for the time being.
The EC asked OICA-CLEPA if there are any specific plans for addressing longitudinal control in specific features.
OICA-CLEPA stated that longitudinal control addresses speed and following distance. For specific aspects of these use cases, elements of longitudinal control might be of interest. OICA-CLEPA suggested to consider longitudinal control as part of an in-depth review of individual features, rather than introducing general functional requirements addressing longitudinal control.
The EC confirmed that they will exchange with OICA-CLEPA and come up with a suggestion.
Addressing 5.6.4.2, OICA-CLEPA proposed to remove reference to emergency interventions at this time due to the wide variety of emergency systems which may have unintended consequences. OICA-CLEPA explained that system deactivation may not be desirable (e.g. in the context of sudden braking due to the close cut-in).
The EC indicated these changes to be acceptable, but expressed that DCAS should remain turned off if it deactivated itself at least until the system is reactivated by the driver.
Japan indicated that DCAS should remain deactivated following an emergency intervention. Japan indicated to prefer the original text by JRC. Japan will provide a proposal explaining emergency based on the JRC proposal.
FIA echoed Japan’s concerns that it should better understood what determines an emergency situation.
The UK questioned why emergency interventions should differentiated from other automatic deactivations by the system. The UK agreed with the point that an intervention from AEBS may not necessarily justify the unavailability of the system which may be confusing for the driver. Alternatively, if the system is outside the operational domain this may make more sense.
ITU questioned whether DCAS should be deactivated if there is an emergency braking, as this is a part of DCAS.
OICA-CLEPA commented that the principle is being established that when the system is automatically deactivated, it cannot be activated without driver action.
Japan suggested to include “DCAS shall be deactivated automatically if emergency systems such as AEBS, RMF and ESF are activated.”.
In response to Japan, OICA-CLEPA questioned whether we were certain whether there were not scenarios where it would be beneficial to maintain continued safe control by DCAS after an emergency intervention. OICA-CLEPA believed a general prohibition would be inappropriate.
ITU concurred with OICA-CLEPA.
On 5.6.5.2, OICA-CLEPA explained that like warnings should be useable in the context of escalation as this might otherwise mean that similar visual cues (such as ‘put your hands on the steering wheel’) may not be used.
The EC agreed with the proposal of OICA-CLEPA.
The Chair informed the TF that the remaining issues will be discussed in the SDG but invited comments to the proposals.
