2nd ACSF Informal meeting
Date: June, 15th – 17th, 2015 Venue: JASIC Tokyo Office
http://www.jasic.org/e/index_e.htm 2.5 days : ACSF Informal Meeting
(From 15th to 17th afternoon) 0.5 days : LKAS Ad-hoc Meeting
(From 17th afternoon)
Access to JASIC Office
Access to JASIC Office
Yotsuya-sanchome Sta. can access from Tokyo sta.(Central Sta.) with Tokyo Metro Marunouchi line. Yotsuya Sta. can access from Tokyo sta. (Central Sta.) with JR Chuo line.
Hotel Name (En) Hotel name (Jn) URL Access to our office H o t e l W I N G International
H o t e l W I N G International
http://www.hotelwingjapan.com/ About 5 min. on foot
MITSUI GARDEN Hotel 三井ガーデンホテル四谷 https://secure.reservation.jp/gardenhotels/stay_pc/rsv/index.aspx?hi_id=3&lang=en-US
About 8 min. on foot
TOKYU STAY YOTSUYA 東急ステイ四谷 http://www.tokyustay.co.jp/e/hotel/YOT/ About 8 min. on foot
APA Hotel (Shinjuku Gyoemmae)
APAホテル<新宿御苑前>
http://www.apahotel.com.e.ju.hp.transer.com/language/shutoken/33_shinjukugyoenmae.html
About 10 min. by Tokyo Metro Marunouchi Line (Yotsuya Sanchome Station)
EISHINKAN 栄進館 http://travel.rakuten.co.jp/HOTEL/936/936.html About 10 min. on foot
Hotel NEW SHOHEI ホテルニューショーヘイ http://www.shoheikan.co.jp/Englishindex.html About 10 min. on foot
Shinjuku CITY Hotel N.U.T.S Tokyo
新宿 CITY HOTEL N. U. T. S 東京
h t t p : / / h o t e l -nuts.com/eng/index.php
About 10 min. by Tokyo Metro Marunouchi Line(Yotsuya Sanchome Station)
TOKYO DOME Hotel 東京ドームホテル h t t p : / / w w w . t o k y o d o m e -hotels.co.jp/e/
About 15 min. by JR Sobu line (Yotsuya station)
If you have any question, please contact with
Mr. Tsuburai
How to access JASIC Office How to access Hotel
ACSF-02-02
DRAFT AGENDA
2nd meeting of the GRRF informal working group on
Automatically Commanded Steering Function (ACSF) Venues: Japan Automobile Standard Internationalization Center (JASIC) Office,
7th floor, Zennihon Truck Sogo Kaikan., 3-2-5 Yotsuya, Shinjuku-ku, Tokyo 160-0004, JAPAN
Chairman: Mr. Christian Theis (Germany) and Mr. Hidenobu Kubota (Japan) Secretariat: Mr. Jochen Schaefer (CLEPA) Duration of the sessions: Tuesday, 16 June 2015: starting at 9 am until Wednesday, 17 June 2015: ending at 4 pm Note: Any comments or documents relating to this meeting should be sent to the secretariat
([email protected]) in e-format, so that meeting documents can be made available to the UNECE secretariat for publication on the website of WP29.
1. Welcome and Introduction
2. Approval of the agenda
Document: ACSF-02-01 (Chair)
3. Addoption of the report of the 1st meeting of the ACSF IG Document: ACSF-01-14 (Secretariat)
4. Confirmation of the homework
Document: ACSF-01-13 (Chair and Secretariat) Some documents will be provided by CPs and Industries.
5. Discussion for draft proposal to GRRF
Document: Please look on the ACSF-website on a frequently basis
6. Confirmation of TOR (Preparation of formal document for next GRRF) and status report for next GRRF
7. Other business
8. List of action items:
Define next steps and list of action items. 9. Schedule for further meetings.
Information about the 3rd/4th meeting (3rd meeting 14th September 2015 in Geneva)
Submitted by the expert from Germany
Proposal for amendments to Regulation No. 79 to include ACSF > 10 km/h
The modifications to the Regulation are marked in blue bold and strikethrough characters.
==================================================================================
Amend paragraph 2.3.4.1. to read:
2.3.4.1. "Automatically commanded steering function" (ACSF) means the function within a
complex electronic control system where actuation of the steering system can result
from automatic evaluation of signals initiated on-board the vehicle, possibly in
conjunction with passive infrastructure features, to generate continuous control
action in order to assist the driver in following a particular path, in low speed
manoeuvring or parking operations.
2.3.4.1.1. Category 1 ACSF means, a function that operates at a speed no greater than 10
km/h to assist the driver, on demand, in low speed manoeuvring or parking
operations.
2.3.4.1.2. Category 2 ACSF means, a function that operates at a speed no greater than [130
km/h] and which can perform a single manoeuver (e.g. lane change) when
commanded by the driver.
2.3.4.1.3. Category 3 ACSF means, a function that operates at a speed no greater than [130
km/h] and which can indicate the possibility of a single manoeuvre (e.g. lane
change) but performs that function only following a command by the driver.
2.3.4.1.3. Category 4 ACSF means, a function that operates at a speed no greater than [130
km/h], which is commanded by the driver and which can continuously determine
the possibility of a manoeuvre (e.g. lane change) and complete these manoeuvers
for extended periods without further driver command.
Insert new paragraph 2.4.8. to read
2.4.8. For Automatically commanded steering functions
2.4.8.1 “Motorway” means, a road section, dedicated exclusively to motor vehicles, having at least
two traffic lanes for each direction of travel and having a physical separation of traffic
moving in opposite directions.
2.4.8.2 "Lane" means one of the longitudinal strips into which a roadway is divided.
2.4.8.3 "Visible Lane markings" means delineators intentionally placed on the borderline of the
lane that are directly visible by the driver while driving (e.g. not covered by snow, etc.).
Informal Document ACSF-02-03
2.4.8.4 "Lead vehicle" means a vehicle driving in front of the vehicle equipped with ACSF.
2.4.8.5 "Lane change manoeuvre" means a manoeuvre in which the vehicle changes from its initial
travel lane to an adjacent lane
2.4.8.6 "Specified maximum speed Vsmax " means the maximum speed up to which an ACSF is
designed to work.
2.4.8.7 "Specified minimum speed Vsmin" means the minimum speed up to which an ACSF is
designed to work.
2.4.8.8 "Transition request" means a request to the driver that the driver has to take over manual
control of the steering task again.
2.4.8.9 "Transition procedure" means the sequence of providing a transition request by the ACSF,
taking over manual steering control by the driver and deactivation of the ACSF since
manual control was detected by the ACSF.
2.4.8.10 "Conditions for safe operation" mean all circumstances like traffic situation, road
category,quality of lane markings, vehicle speed, curvature of the road, lighting, sensor
capacities etc. specified by the vehicle manufacturer that have to be fulfilled when an ACSF
shall be able to be activated by a driver.
2.4.8.11 "System boundaries" mean all circumstances from which on the conditions for safe
operation are not fulfilled anymore, that cannot be dealt with by an activated ACSF
anymore and thus request a take-over of manual steering control by the driver.
2.4.8.12 "ACSF status" means any distinct operational mode of the ACSF like "switched off"
"switched on", "available to be activated", "activated" etc.
2.4.8.13 "Attention recognition system" means a device to detect if the driver is vigilant, is
attentive, is aware of the traffic situation
2.4.8.14 "Minimum risk manoeuvre" means a strategy for the [longitudinal and] lateral control of
the vehicle to reach a status with as little risk as possible in the given traffic situation when
the driver is detected by the ACSF not to be available.
Amend paragraph 5.1.6.1. to read:
5.1.6.1. Whenever the an Automatically Commanded Steering function becomes operational, this
shall be indicated to the driver. and the control action shall be automatically disabled if the
vehicle speed exceeds the set limit of 10 km/h by more than 20 per cent or the signals to be
evaluated are no longer being received. Any termination of control shall produce a short but
distinctive driver warning in accordance with the requirements of paragraph 5.4.3. by a
visual signal and either an acoustic signal or by imposing a tactile warning signal on the
steering control.
Insert new paragraph 5.4.3. Renumber paragraph 5.4.3. as 5.4.4.
5.4.3. Special Warning Provisions for Automatically Commanded Steering Functions
5.4.3.1 Any termination of control shall produce a distinctive driver warning by a [yellow] visual
signal and either an acoustic signal or by imposing a haptic warning signal. This warning
shall be provided before the system (function) becomes in-operational, if the termination is
not intended by the driver.
5.4.3.2. Any sudden termination of control caused by a failure of the system physical or functional
failure shall produce immediately a short but distinctive driver warning by a [red] visual
signal and either an acoustic signal that shall remain operational until the driver has
resumed control.
Insert new paragraph 5.6
5.6 Special Provisions for Automatically Commanded Steering Functions
5.6.1. Special Provisions for ACSF of Category 4
5.6.1.1. General
5.6.1.1.1. The vehicle shall be equipped with a means for the driver to activate or deactivate
the ACSF at any time.
5.6.1.1.2. The ACSF shall be activatable only if the conditions for safe operation of the ACSF
are fulfilled [all associated functions – brakes, accelerator, steering,
camera/radar/lidar etc. are working proper).
5.6.1.1.3. The ACSF shall only activate by deliberate action of the driver.
5.6.1.1.4. The ACSF shall be able to detect if the driver controls the steering function
manually. If the ACSF detects, that the driver is steering manually , ACSF shall be
deactivated.
5.6.1.1.5. The ACSF may be operational up to a vehicle lateral acceleration of [3] m/s2.
5.6.1.1.6. The ACSF shall comprise an attention recognition system that is active whenever
the ACSF is active.
5.6.1.2. Operation of ACSF
5.6.1.2.1. Any lane change manoeuvre shall be initiated only if:
- the vehicle is travelling on motorway as defined in paragraph 2.4.8. and
- any traffic that can affect the safe manoeuvre shall be identified by equipment
installed on the vehicle and
- the vehicle equipment can analyse speed and distance of the identified traffic to
ensure a safe manoeuvre(e.g. does not cause a deviation to the flow or direction of
other traffic).
5.6.1.2.2. If a lane change manoeuvre is carried out, the correspondent direction indicator
lamps shall be automatically activated.
5.6.1.2.3. If the ACSF is not overridden by the driver it shall not terminate the lane change
until the manoeuvre is safely completed, exept for the ACSF detects an imminent
critical situation.
5.6.1.2.4. The activated ACSF shall at any time control the lateral movements of the vehicle in
such a way that the vehicle does not induce any safety critical situations and that
the movements of the vehicle are clear to other road users.
5.6.1.2.5. The activated ACSF shall at any time ensure a safe lateral distance to other road
users. The vehicle manufacturer shall provide documentation about how such a
safe distance is achieved to the technical service.
5.6.1.2.6. The ACSF shall be designed such that safe transition to manual steering is possible
at any time.
5.6.1.2.7. If the activated ACSF detects that due to a sudden unexpected event the vehicle is
in imminent danger to collide with another road user and that the time for a safe
transition procedure is too short, an emergency manoeuvre shall be carried out.
The vehicle manufacturer shall provide information to the technical service about
the safety strategy depending on different circumstances forming a sudden critical
event and the foreseen emergency manoeuvres.
5.6.1.2.8. If the attention recognition system detects that the driver is inattentive, it shall give
a warning to restore attentiveness again. The manufacturer shall provide
information to the technical service how the attention recognition systems detects
inattentiveness of the driver.
5.6.1.3. System boundaries
5.6.1.3.1. The vehicle manufacturer shall provide the values for Vsmax and Vsmin to the technical
service.
5.6.1.3.2. The vehicle manufacturer shall provide an information to the technical service
under which conditions an ACSF can be activated, i. e. when the conditions for safe
operation of the ACSF are fulfilled.
5.6.1.3.3. The vehicle manufacturer shall provide information to the technical service about
system boundaries at which the activated ACSF must give a transition request.
5.6.1.4. Indication of ACSF status
5.6.1.4.1. The ACSF shall at any time give a noticable and distinctive signalisation to the driver
about the ACSF status. This signalisation shall be at least a visual signal. Any change
in system status shall be indicated by an optical and either an acoustic or haptic
signal.
5.6.1.5. Transition request
5.6.1.5.1. If ACSF detects that its system boundaries are reached or will be reached shortly it
shall provide a transition request.
5.6.1.5.2. The timing of the transition request shall be such that sufficient time is provided for
a safe transition of the steering task from automatically commanded steering to
manual steering.
5.6.1.5.3. The vehicle manufacturer shall provide specific values for time intervals to the
technical service, which are foreseen for safe transition under different
circumstances.
5.6.1.5.4. If the speed of the vehicle with activated ACSF exceeds vsmax a transition request
shall be given.
5.6.1.5.5. If the vehicle reaches a lateral acceleration of more than [3] m/s2 a transition
request shall be given.
5.6.1.5.6. If an attention recognition system detects the driver to be inattentive although a
warning to restore attentiveness was provided to the driver a transition request
shall be given.
5.6.1.5.7. The ACSF shall provide a transition request if the driver's seatbelt is unfastened
and/or if the driver's seat is left by the driver.
5.6.1.5.8. The transition request shall be provided by a [yellow] visual signal and either an
acoustic signal or by imposing a haptic warning signal.
5.6.1.5.9. If the driver does not take over manual control immediately the transition request
shall be escalating with time in terms of enlarging the intensity of the warning
and/or in terms of adding and/or changing the warning means.
5.6.1.5. Minimum Risk Manoeuvre
5.6.1.5.1. If the ACSF detects that after a transition request the driver does not take over
manual control of the steering again the vehicle shall carry out a minimum risk
manoeuvre. The vehicle manufacturer shall provide information to the technical
service about which kind of minimum risk manoeuvres are foreseen depending on
the given traffic situation and circumstances at its initiation.
5.6.2. Special Provisions for ACSF of Category 3
- to be developed based on the requirements for a Category 4 system-
.
.
5.6.3. Special Provisions for ACSF of Category 2
- to be developed based on the requirements for a Category 4 system-
.
.
5.6.4. Special Provisions for ACSF of Category 1
- to be developed based on the requirements for a Category 4 system-
.
Insert new Annex 7
Annex 7
Text requirements for automatically commanded steering functions
1. General Provisions
Vehicles fitted with ACSF shall fulfill the tests requirements of this annex according to the
corresponding category of ACSF specified in Table 1.
2. Test conditions
2.1. The test shall be performed on a flat, dry asphalt or concrete surface delivering good adhesion. The ambient temperature shall be between 0° C and 45° C.
2.2. Lane markings
The lane markings and the width of the lane used in the tests shall be those of one of the Contracting Parties, with the markings being in good condition and of a material conforming to the standard for visible lane markings of that Contracting Party. The lane marking layout used for the testing shall be recorded.
The test shall be performed under visibility conditions that allow safe driving at the required test speed.
2.3 Lead vehicle
The lead vehicle shall be a high volume series production passenger car of category M1 AA saloon or in the case of a soft target an object representative of such a vehicle in terms of its detection characteristics. A soft target is a target that will suffer minimum damage and cause minimum damage to the subject vehicle in the event of a collision.
2.4 Pedestrian soft target
A pedestrian soft target is an object representative of a human adult in terms of its detection characteristics that will suffer minimum damage and cause minimum damage to the subject vehicle in the event of a collision.
3. Tests
Table 1 specifies which tests have to be fulfilled by each ASCF category.
Test \ ACSF category 1 2 3 4
FU1 path following with lead X
FU2 lane keeping w/o lead X
TR1 tight curve with lead X
TR2 tight curve w/o lead X
TR3 no marking with lead X
TR4 no marking w/o lead X
EM1 AEB on braking rabbit X
EM2 AEB on static pedestrian X
EM3 abortion of lane change X
Table 1
3.1. Functionality Tests
3.1.1. Functionality Test 1 (FU1)
Drive the vehicle with activated ACSF at least 5 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with various
curvatures with road markings at each side of the lane at various speeds up to vsmax and
down to vsmin. The speed of the lead vehicle shall be selected such that the lateral
acceleration is not more than 1 m/s².
The requirements of the test are fulfilled if the vehicle does not cross any lane marking.
3.1.2. Functionality Test 2 (FU2)
Drive the vehicle with activated ACSF at least 5 min without a lead vehicle. The vehicle shall
drive on a track with various curvatures with road markings at each side of the lane at
various speeds up to vsmax and down to vsmin. The speed shall be selected such that the
lateral acceleration is not more than 1 m/s².
The requirements of the test are fulfilled if the vehicle does not cross any lane marking.
3.2. Transition Tests
3.2.1 Transition Test 1 (TR1)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with road markings
at each side of the lane at a speed of 10 km/h below vsmax. After a straight section of at
least 200 m the lead vehicle shall enter a curve of more than 90° that demands a lateral
vehicle acceleration of more than 3 m/s2. The test driver of the vehicle shall not take over
manual steering control again.
The requirements of the test are fulfilled if the transition request was given at least when
the lateral acceleration exeeds [3] m/s² and the minimum risk maneuvre as specified by
the manufacturer was initiated. The vehicle shall not cross any lane marking before the
minimum risk maneuvre was initiated.
3.2.2. Transition Test 2 (TR2)
Drive the vehicle with activated ACSF at least 1 min without a lead vehicle. The vehicle shall
drive on a track with road markings at each side of the lane at at a speed of 10 km/h below
vsmax. After a straight section of at least 200 m the vehicle shall approach a curve of more
than 90° that would demand a lateral vehicle acceleration of more than 3 m/s2. The test
driver of the vehicle shall not take over manual steering control again.
The requirements of the test are fulfilled if the transition request was given at least when
the lateral acceleration exeeds [3] m/s² and the minimum risk maneuvre as specified by
the manufacturer was initiated. The vehicle shall not cross any lane marking before the
minimum risk maneuvre was initiated.
3.2.3. Transition Test 3 (TR3)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with road markings
at each side of the lane at a speed of 10 km/h below vsmax. After a straight section of at
least 200 m the lead vehicle shall enter a section with a length of 200 m with only one lane
marking at the driver‘s side. The test driver of the vehicle shall not take over manual
steering control again.
The requirements of the test are fulfilled if the vehicle does not cross any lane marking and
if the transition request was given [0 s] before the vehicle would have entered the section
with missing lane marking and the minimum risk maneuvre as specified by the
manufacturer was initiated.
3.2.4 Transition Test 4 (TR4)
Drive the vehicle with activated ACSF at least 1 min without a lead vehicle. The vehicle shall
drive on a track with road markings at each side of the lane at at a speed of 10 km/h below
vsmax. After a straight section of at least 200 m the vehicle shall approach a section with a
length of 200 m with only one lane marking at the driver‘s side. The test driver of the
vehicle shall not take over manual steering control again.
The requirements of the test are fulfilled if the vehicle does not cross any lane marking and
if the transition request was given [5 s] before the vehicle would have entered the section
with missing lane marking and the minimum risk maneuvre as specified by the
manufacturer was initiated.
3.3. Emergency Tests
3.3.1 Emergency Test 1 (EM1)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the system, the vehicle shall drive at a gap of 3 s behind the lead vehicle.
The lead vehicle shall drive within the lane markings on a track with road markings at each
side of the lane at a speed 10 km/h below vsmax. Then the lead vehicle decelerates with 6
m/s² and with a mean brake jerk of 6 m/s³ in the first second of braking.
The requirements of the test are fulfilled if the vehicle does not collide with the lead
vehicle.
3.3.2. Emergency Test 2 (EM2)
Drive the vehicle with activated ACSF at least 1 min on a track with road markings at each
side of the lane at a speed 10 km/h below vsmax. The vehicle shall approach a stationary
pedestrian soft target being placed in the center of the lane.
The requirements of the test are fulfilled if the vehicle does not collide with the pedestrian
soft target.
3.3.3. Emergency Test 3 (EM3)
Drive the vehicle with activated ACSF at least 1 min on a track with road markings at each
side of the lane at a speed of [30 km/h] below vsmax or at vsmin, whatever is higher, between
two other vehicles. If the time gap is not selected by the system, the vehicle shall drive at a
gap of 3 s behind the lead vehicle. Induce by e.g. selecting a higher desired speed a lane
change manoeuvre. During the lane change a further overtaking vehicle with a speed of 50
km/h above vsmax shall approach from backwards on the adjacent lane that was intended to
be reached with the lane change. At the point in time when the vehicle is crossing the lane
marking the further overtaking vehicle shall be at a TTC of [2 s] behind the vehicle.
The requirements of the test are fulfilled if the vehicle does not collide with the further
overtaking vehicle or any other vehicle and aborts the lane change manoeuvre.
1
Submitted by the expert from Germany
HMI concept of ACSF - background knowledge from research
The development of automatically commanded steering functions (ACSF) presents
different challenges for HMI design in vehicles. In comparison to manual steering or
to some established low-functional driver assistant systems, the main focus lies on
action by the human driver who is - with regard to the steering task - only sitting in the
vehicle (not steering). As mentioned in many research papers (e.g. Bengler & Flem-
isch, 2011) as well as in Gasser et al. (2012), the driving task changes increasingly
from a controlling and regulating procedure to a monitoring procedure. This requires
an increase in human cognitive workload (Endsley & Kiris, 1995) and permanent at-
tention (vigilance) because of potential transitory demands. Regarding this vigilance,
the HMI design has to tackle the crucial challenge of maintaining human attention
(see Parasuraman, Mouloua & Molloy, 1996; Muhrer & Vollrath, 2011; Vollrath,
Schleicher & Gelau, 2011; Neubauer, Matthews, Langheim & Saxby, 2012).
Especially for automated steering systems, it is necessary to follow basic aspects of
human sensation and perception in the automotive context, as well as to verify the
transferability of well-known HMI design principles of driving assistant systems
(ESoP, 2006; DIN EN ISO 9241-110). Specifically, aspects like compatibility, con-
sistency, configuration in space, balance between mental underload and mental
overload, comfort and a holistic view of the HMI have been proven to be effective
(see Bruder & Didier, 2015). One of the most important issues is explicit information
about the system mode to prevent mode-confusion (Bengler & Flemisch, 2011).
In general, the HMI must be able to get the driver safely back into the loop again and
provide him with adequate situational awareness (Merat & Jamson, 2009; Vollrath &
Krems, 2011) after he has merely monitored the driving procedure. Buld and Krüger
(2003) as well as Muhrer and Vollrath (2011) report higher collision rates in a moni-
toring driving task compared to manual driving. Furthermore, people often do not de-
tect system errors when driving just by monitoring (Niederée & Vollrath, 2009). At
Informal Document ACSF-02-04
2
last, drivers’ reaction times (in this case speed reduction) are much longer in a moni-
toring driving task and differ in about five seconds to manual driving
(Vollrath et al., 2011).
Current research indicates that warnings have a positive effect (Merat & Jamson,
2009; Dogan, Deborne, Delhomme, Kemeny & Jonville, 2014), because they reduce
drivers’ reaction times compared to situations without a warning (Fricke & De Flippis,
2008; Lee, McGehee, Brown & Reyes, 2002; Flemisch et al., 2011). By using a com-
bination of visual and acoustic warnings instead of only visual Naujoks, Mai and
Neukum (2014) achieved better driver reaction in transitory situations. Therefore it is
important to create redundancies between all warning options (visual, acoustical and
tactile) in line with Wickens’ Multiple Resource Theory (Wickens, 2008). Comparing
hands-on and hands-off driving in different automatic scenarios, first results (Gold,
Lorenz, Damböck & Bengler, 2013) tend to indicate faster driver intervention with
hands-on. It must be noted that all current results provide some indications for spe-
cial driving and transitory situations in their respective testing scenario only: Many
factors play an influencing and moderating role in finding a universal solution for HMI
design for automated driving. Apart from the driver’s condition the driving situation is
another important factor, as Kleen and Vollrath (2012) have shown. Moreover, it is
necessary to consider people’s experience with established driver assistant systems
(Weinberger, Winner & Bubb, 2001) when thinking about HMI design of automated
steering systems, because learning of the system functions and its limits may take
place (Strand, Nilsson, Karlsson & Nilsson, 2014).
In general – considering current research – there are merely some indications and
tendencies for HMI design for ACSF with limitations of transferability and validity, in-
volving driving simulation. Most questions and problems with regard to the HMI for
automated driving tasks have still to be answered and solved in further research in
national as well as in international projects.
3
References
Bengler, K., & Flemisch, F. (2011). Von H-Mode zur kooperativen Fahrzeugführung–
grundlegende ergonomische Fragestellungen. In 5. Darmstädter Kolloquium – Zu-
kunft der Fahrzeugführung. Kooperativ oder autonom.
Bruder, R., & Didier, M. (2015). Gestaltung von Mensch-Maschine-Schnittstellen. In:
Handbuch Fahrerassistenzsysteme, 3. Auflage. Springer Vieweg.
Buld, S., & Kruger, H. (2003). Die Auswirkung von Teilautomation auf das Fahrver-
halten. DGLR BERICHT, 4, 241.
DIN EN ISO 9241-110: Ergonomie der Mensch-Maschine-Interaktion. Teil 110:
Grundsätze der Dialoggestaltung, 2006.
Dogan, E., Deborne, R., Delhomme, P., Kemeny, A., Jonville, P. (2014). Evaluating
the shift of control between driver and vehicle at high automation at low speed: The
role of anticipation. Transport Research Arena 2014. Paris, France.
ESoP (2006): COMMISSION RECOMMENDATION of 22 December 2006 on safe
and efficient in-vehicle information and communication systems: update of the Euro-
pean Statement of Principles on human machine interface (2007/78/EC), L 32/200,
Official Journal of the European Union, 6.2.2007
Endsley, M. R., & Kiris, E. O. (1995). The out-of-the-loop performance problem and
level of control in automation. Human Factors: The Journal of the Human Factors and
Ergonomics Society, 37(2), 381-394.
Flemisch, F., Schieben, A., Schoemig, N., Strauss, M., Lueke, S., & Heyden, A.
(2011). Design of human computer interfaces for highly automated vehicles in the
EU-Project HAVEit. In Universal Access in Human-Computer Interaction. Context
Diversity (pp. 270-279). Springer Berlin Heidelberg.
4
Fricke, N. & De Filippis, M. (2008). Effects of auditory warnings on driving behaviour.
In: D. de Waard, F. O. Flemisch, B. Lorenz, H. Oberheid, & K. A. Brookhuis (Hrsg.),
Human Factors for assistance and automation (pp. 117-128). Maastricht, the Nether-
lands: Shaker Publishing.
Gasser, T.M., Arzt, C., Ayoubi, M., Bartels, A., Bürkle, L., Eier, J., Flemisch, F.,
Häcker, D., Hesse, T., Huber, W., Lotz, C., Maurer, M., Ruth-Schumacher, R.,
Schwarz, J., Vogt, W. (2012). Rechtsfolgen zunehmender Fahrzeugautomatisierung.
Berichte der Bundesanstalt für Straßenwesen. Fahrzeugtechnik, vol. 83. Wirtschafts-
verlag NW, Bergisch Gladbach.
Gold, C., Lorenz, L., Damböck, D., Bengler, K. (2013). Partially Automated Driving as
a Fallback Level of High Automation. 6. Tagung Fahrerassistenzsysteme. Der Weg
zum automatischen Fahren. TÜV SÜD Akademie GmbH.
Kleen, A. & Vollrath, M. (2012). Beherrschbarkeit von komplexen Eingriffen in die
Fahrzeugführung. In: VDI Fahrzeug- und Verkehrstechnik (Hrsg.). Fahrerassistenz
und Integrierte Sicherheit. 27. VDI/VW-Gemeinschaftstagung. Düsseldorf: VDI Verlag
GmbH.
Lee, J. D., McGehee, D. V., Brown, T. L., & Reyes, M. L. (2002). Driver distraction,
warning algorithm parameters, and driver response to imminent rear-end collisions in
a high-fidelity driving simulator. Human Factors, 44, 314-334.
Merat, N. & Jamson, A. H. (2009). How do drivers behave in a highly automated car?
In: M. Rizzo, J.D. Lee, & D. V. McGehee (Hrsg.). Proceedings: The Fifth Annual In-
ternational Driving Symposium On Human Factors In Driver Assessment, Training
And Vehicle Design (pp. 514-521). Iowa City, IA: The University of Iowa.
Muhrer, E., & Vollrath, M. (2011). Das Projekt Isi-Padas - Ein Überblick. In VDI-
Verlag (Ed.)6, 6. VDI-Tagung Der Fahrer im 21. Jahrhundert (pp. 207-221). Düssel-
dorf: VDI.
5
Naujoks, F., Mai, C., & Neukum, A. (2014). The effect of urgency take-over requests
during highly automated driving under distraction conditions. In: T. Ahram, W. Kar-
wowski, & T. Marek (Hrsg.). Proceedings of the 5th International Conference on Ap-
plied Human Factors and Ergonomics AHFE 2014. Krakow, Polen.
Neubauer, C., Matthews, G., Langheim, L., & Saxby, D. (2012). Fatigue and Volun-
tary Utilization of Automation in Simulated Driving. The Journal of the Human Factors
and Ergonomics Society, 54(5), 734–746.
Niederée, U., & Vollrath, M. (2009). Fahrerassistenzsysteme der Zukunft - Fährt da
der Mensch noch mit?. VDI-Berichte, (2085).
Parasuraman, R., Mouloua, M., & Molloy, R. (1996). Effects of adaptive task alloca-
tion on monitoring of automated systems. Human Factors, 38(4), 665-679.
Strand, N., Nilsson, J., Karlsson, I. C. M., Nilsson, L. (2014). Semi-automated versus
highly automated driving in critical situations caused by automation failures. Trans-
portation Research Part F: Traffic Psychology and Behaviour, 27, 218-228.
Vollrath, M. & Krems, J. (2011). Verkehrspsychologie. Ein Lehrbuch für Psychologen,
Ingenieure und Informatiker. Kohlhammer Standards Psychologie.
Vollrath, M., Schleicher, S., & Gelau, C. (2011). The Influence of Cruise Control and
Adaptive Cruise Control on Driving Behaviour-A Driving Simulator Study. Accident
Analysis & Prevention, 43(3), 1134-1139.
Weinberger, M., Winner, H., & Bubb, H. (2001). Adaptive cruise control field opera-
tional test - the learning phase. JSAE Review, 22, 487–494.
Wickens, C. D. (2008). Multiple Resources and Mental Workload. Human Factors:
The Journal of the Human Factors and Ergonomics Society, 50(3), 449–455.
Research on HMIHomework item 1 (ACSF-01-13)
ACSF-02-05
Toshiya Hirose, Toru Kojima
Vehicle Safety Research Department
National Traffic Safety and Environment Laboratory, Japan
2nd Meeting of ACSF Informal Group, 15-17 June 2015
Contents of the presentation
1. Report on experiment on behaviors of drowsy drivers during transition from system control to manual driving
2. Research plan for ACSF in 2015
1. Report on experiment on behaviors of drowsy drivers during transition from system control to manual driving
Purpose of the experiment
By using driving simulator,
1.Correcting data on driver’s behaviors during transition from system control to manual driving
2.Comparison of drivers’ behaviors under different level of drowsiness
Audible warning is activated during transition from system control to manual driving
Direction 3.5m
3.5m
Vehicle controlled by a system
Vehicle cutting in
Direction
100km/h
60km/h
Condition and scenario of the experiment
Subject : 40 test drivers (20-30 years old) Scenario: The driver is sitting on the Driving Simulator which simulate driving at
100 km/h on a high way. Another vehicle cuts in at 60 km/h ahead of the vehicle. Status of the drivers: Normal and drowsy Data: Drivers’ operation and their reaction time
Normal
Result of the experiment
Drivers’ operation for collision avoidance
Drowsy
Braking 85%
Analysis of drivers’ behaviors by their reaction time
Level of drowsy Status of drivers Physical signs
1 Not sleepy at all Blink a stable
2 Not sleepy Yawn
3 Sleepy Re-sit
4 Sleepy fairly Shaking a head
5 Sleepy very Close the eyelid
Level of drowsy
Definition of reaction time
The time from the moment the forward vehicle cutting in the same lane to the timing of braking operation by the driver
Reaction time: 0.3-2.5 secBrake Pedal force: 70-240 N
Reaction time and brake pedal force by normal drivers
Level of drowsy
Reaction time: 2.0-4.0 secBrake Pedal force: 50-230 N
Reaction time and brake pedal force by drowsy drivers
Level of drowsy
Normal: 1.78sec, Drowsy: 2.66secReaction of drowsy drivers is longer by 0.9 seconds, or 50% than that of normal drivers
Comparison of reaction time (normal vs drowsy drivers)
Drowsy Normal
Reference: SAE Technical Paper 2015-01-1407, 2015, doi:10.4271/2015-01-1407
2. Research plan for ACSF in 2015
Assessment of the drivers’ behaviors during transition from ACSF to manual driving. (drivers’ operation, reaction time, etc.)Several scenarios (e.g. malfunction) are considered in the test.
!Malfunction
Lane Keeping
Examples of the traffic scene (1)(malfunction during lane keeping on a curve )
• Notification timing from HMI to the driver1)[4s] before the malfunction2)[2s] before the malfunction]3)Same timing as the malfunction
Steering Torque by ACSF
Time
Malfunction
type 1
type 2
type 3
Drivers’ behaviors during the transition from ACSF to manual driving under several scenarios
Examples of the traffic scene (2)(Malfunction during a lane change)
Examples of the traffic scene (3)(Collision avoidance with other vehicle at the junction)
Malfunction
Lane Change (Overtaking)
!
Lane Change (Overtaking)
Override by the driver
• Lateral velocity during lane change1)[0.2~0.4] m/s
• Notification timing from HMI to the driver
1)[TTC 4s]2)No notification
• Lateral velocity during lane change1)[0.2~0.4] m/s2)[0.6~0.8] m/s3)[1.0~1.2] m/s
• Notification timing from HMI to the driver
1)[4s] before the malfunction2)[2s] before the malfunction3)Same timing as the malfunction
Examples of the traffic scene (4) (reduction of the lanes)
Sub-task for the driver during using ACSF (applied to all experimental traffic scenes)
Lane Keep
•Notification timing from HMI to the driver
1)[TTC 4s]2)No notification
Override by the driver
1)Mental calculation (the driver sees the front, but his/her attentiveness will be decreased)
2)Operation of a smart phone or drinking coffee etc. (the driver doesn’t see the front
sufficiently)
3)No sub-task (the driver always sees the front)
Research schedules for ACSF in 2015
By the end of August• To modify the driving simulator.• To make experimental scenarios
By the end of September• To conduct experiment
By the end of October• To analyse data
OICA/CLEPA Homework Part IHMI, Driver in/out of the Loop
Submitted by the experts of OICA/CLEPA
Tokyo, 16-17 June 2015
Informal Document ACSF-02-06
Contents
• Examples for ACSF Use Cases
• ACSF Control Loop
• HMI Communication between ACSF and the Driver– HMI Information and Warning
– Transition to/from Manual Driving
– Evaluation of Driver‘s Activity
• Being active vs. being attentive
• Scenarios for Active/Attentive Driving
• Driver in/out of the Loop (DIL/DOL)
ManeuverStep
DescriptionLK LK & LC
LK & LC automated
1 ACSF manually activated by the driver X X X
2 Continuous detection of the environment front/side/rear
x/-/- x/x/x x/x/x
3 ACSF System Activation (displayed to the driver ) X X X
4 Lane Keeping Functionality active X X X
5 Manual Activation of LC (e.g. turn indicator) X
6 Lane Change Readiness X X
7 Start and performing Lane Change ( change of steering angle)
X X
8 Arrived in new lane X X X
9 Lane Keeping X X X
10 ACSF issues takeover request to the driver X X X
11 Driver takes over the driving taskACSF is deactivated
Examples For ACSF Use Case Analysis
Dri
ver
can
ove
rste
erth
eA
CSF
at
any
tim
e
ACSF Control Loop
Driver
HMI
Oversteering of ACSF at any time
-A
ctiv
ate
/De
acti
vate
-Ev
alu
atio
n o
fd
rive
r‘s
ph
ysic
alac
tivi
ty
The HMI is part of the "ACSF control loop“. In order to better structure the HMI discussion, 3 sub-categories are proposed (with focus on technology neutral solutions and design freedom):
• Information and warning about ACSF status- Information to the driver on ACSF operational status- Warning on termination/sudden termination of ACSF control
• Evaluation of the driver’s activity (suggested in the Proposal by D)- Means to detect the presence and activity of the driver (in combination with a warning strategy)
• Transition to/from manual control- Provide means to activate/deactivate the ACSF operation- Detect manual control/takeover by the driver - Ensure safe transition to/from manual control
HMI - Communication between ACSF and the Driver
HMI – Information and WarningCommunication between ACSF and the Driver
• The design options of communicating with the driver are manifold, as investigatedby the HAVEit* Research Project
• The communication modalities can be designed by visual, acoustic and hapticmeans and by any combination (open for new technologies)
• Each modality can be further classified, as indicated in the table below
*Source: HAVEit (Highly Automated Vehicles for Intelligent Transport)Preliminary concept on optimum task repartition for HAVEit systems
Detect safely when driver • has ceded lateral control to the system and inform the driver• has taken back manual control
Different transition scenarios • Time-critical emergency transition (e.g. sensor failure)• Driver initiated transition
Sensing of steering and/or pedal activity
Transition to/from Manual Control
• Detect presence of the driver
• When ACSF is active, the driver’s role is to remain attentive and to be aware of the system’s/vehicle’s status. The status of the system shall be displayed in the direct view of the driver. The driver shall remain in a position to be able to resume the dynamic driving task whenever required. The system has to provide pre-requisites for the evaluation of the driver’s physical activities.
Evaluation of Driver‘s Activity
Scenarios for Active/Attentive Driving
Driver is active Driver is inactive
Driver is attentive Normal driving Happens for short period of time on a straight road during normal driving
or
Voluntary misuse, e.g. driver isassessing the driving assistance systems.
Driver is inattentive The driver is keeping the path but is distracted / day-dreaming. This happens for short periods of time during normal driving;
May happen for short period of time during normal driving
or
The driver is fully occupied with something else, voluntarily or not.
1 2
3 4
Driver in/out of the Loop (DIL/DOL)
The R.E.3 Definition of DIL suggests that a driver has to perform the following tasks:i. Perform the longitudinal and lateral dynamicsii. Monitor the driving environmentiii. Be aware of the vehicle status and thus also acknowledge warnings issued by sub-
systems
• Increasing vehicle automation transfers some of the tasks i. – iii. from the driver to thesystem
• Automated driving systems that perform the tasks i.-iii. entirely in a given use-case andhence do not require the presence of a driver are considered as DOL-systems
The prospective automated driving systems require that the driver retains at least oneof the three aforementioned tasks during the entire use-case. Hence, these automateddriving systems remain in the scope of the UN-R 79 regarding steering (ACSF) and arecovered by the Vienna Convention of Road Traffic.
ADAS Principle: R.E.3 Annex5-Appendix 3 (Extract)
Driver in the Loop
The notion of driver-in-the-loop means that a driver is involved in driving task and is aware of the vehicle status and road traffic situation. Being in-the-loop means that the driver plays an active role in the driver-vehicle system. They actively monitor information, detect emerging situations, make decisions and respond as needed.
1
ACSF Test ProcedureDraft proposal – For discussion
OICA and CLEPA proposal for the IG Group ACSFTokyo, 2015, June 16-17
Informal Document ACSF-02-07
• The objective is to present potential test scenario for the ACSF approval, for discussion.
• The detailed test conditions, criterias will be defined once the principleof the tests will be agreed.
• It should also be considered the influence of different types of vehiclesto be approved, e.g. passenger cars, LCVs and HCVs.
Objectives of this document
2
ACSF Test Scenario – General Principles
• Define few key test scenario, with specific test procedure and pass/fail criteria
• Since these test scenarii cannot cover all potential situations in an exhaustive manner, the paragraph 4 « verification and test » of the CEL Annex shall be used.
• The informal group on ACSF deals with the steering aspects of ACSF
3
The basic tests focus on three major capabilities of the system:
• Lane Change under multiple lane traffic conditions – respecting a safe distance
• Curve Travel – ability of the vehicle to follow a curve
• Transition in case of missing lane markings – testing the specified transition time
During these tests, additional basic capabilities are verified, e.g.:
• Detection of objects/lanes
• Keeping the lane
• Keeping the safe distance
ACSF Test Scenario – General Principles
4
ACSF Test Scenario – General PrinciplesSafe Distance Parameters
Safety DistanceShall not be entered by
ACSF vehicleR
R
Safety Distance shall not be entered by ACSF vehicle at any time. Distances are based on the rules
• Safe travelling distance = [2s] at travelling speed
• Typical driver reaction time [1.0s]
• Safe uncritical deceleration [3m/s²]
The lane change maneuver may be simplifyed into two segments of constant lateral acceleration (1m/s²)
For all speed indications, a tolerance of +- 5,0km/h may be applied
Safety DistanceShall not be entered
by ACSF vehicle
ManeuveringDistance
Covered by ACSF vehicle
Safety DistanceShall not be entered
by ACSF vehicle
5
Scenarios of higher complexity may be assesed according to the methodology prescribed by the Annex 6, Complex Electronic Systems, including:
• Safety concept, (to address system integrity and thereby ensure safe operation even in the event of an electrical failure) § 2.1
• Boundary of functional operation, § 2.8
• Safety Analysis“3.4.4. The documentation shall be supported, by an analysis which shows, in overall terms, how the system will behave on the occurrence of any one of those specified faults which will have a bearing on vehicle control performance or safety.This may be based on a Failure Mode and Effect Analysis (FMEA), a Fault Tree Analysis (FTA) or any similar process appropriate to system safety considerations. The chosen analytical approach(es) shall be established and maintained by the Manufacturer”
• Limits defining the boundaries of functional operation (paragraph 2.8.) shall be stated where appropriate to system performance
These potential extra-test shall be conducted following CEL Annex paragraph 4 requirements on “Verification and testing”
ACSF Test Scenario – General Principles
6
›
ACSF Complex Test Scenario – Lane Change
e.g. 80 km/h
e.g. 130 km/h
e.g. 120 km/h
• Safety distances have to be respected• Vehicles on the adjacent lane shall not be forced to brake stronger than 3m/s²• Test condition: e.g. [Vback = 130 km/h], [Vego = 120 km/h], [Vfront = 80 km/h]
• Problem: It is very difficult to synchronize three or more vehicles with respect to speed and distance on a testtrack
7
ACSF Test Scenario – Approaching a Vehicle Test 1: Recognize solid marking
8
e.g. 130 km/h e.g. 100 km/he.g. 100 km/h
Safety Distance
Pass/fail criteria:Cat 2: the ACSF does not change lane even under driver’s requestCat 3: the ACSF does not change lane and shall not propose lane changeCat 4: the ACSF does not change laneNote: For ACSF with only LK capability, the ACSF shall keep lane
Initial test conditions: all vehicle speeds are constant
ACSF Simplified Test Scenario – Lane ChangeTest 2: Recognize occupied lanes in congestion conditions lane change should not be performed
9
V2 = [60 km/h]
[60 km/h]
Vehicle to be recognized by the system
V1 = [80 km/h]
Safe travelling distance
dt = (v2-v1) t [t= 2s] as required in traffic regulation
A short (t <[0,5] s) diving-in into this area is acceptable
Need for specific values for heavy vehicles
Safety Distance
Initial test conditions: all vehicle speeds are constant
ACSF Simplified Test Scenario – Lane ChangeTest 3: Keep safety distance of car behind on adjacent lane lane change should not be performed
10
[80 km/h]
[80 km/h]if d < dsafed
Test conditions: all vehicle speeds are constant
Pass/fail criteria:Cat 2: the ACSF does not change lane even under driver’s requestCat 3: the ACSF does not change lane and shall not propose lane changeCat 4: the ACSF does not change laneNote: For ACSF with only LK capability, the ACSF shall keep lane
ACSF Test Scenario – Following a Curve
11
R=150m
e.g. 60 km/h
Test conditions: vehicle speed is constant
Pass/fail criterion: the vehicle follows the curveQuestion: possible additional test assessing the maximum lateralacceleration permitted for the system?
ACSF Test Scenario – Missing Lane Markings
12
dtrans = v * ttrans
Warning
Test conditions: vehicle speed is constant
ACSF - Traffic Jam Assist on all roads
OICA and CLEPA proposal for the IG Group ACSFTokyo, 2015, June 16-17
Informal Document ACSF-02-08
Traffic Jam Assist on all roads
• HCV manufacturers have a high interest for ACSF at “cruise speed” (90km/h), but also for ACSF in traffic jam conditions at lower speeds (e.g. below 40km/h).
• We agree it is a reasonable way to start implementing in a limited environment, i.e. on highway with constructional separation between the two traffic directions.
• However, this condition should only apply to “cruise speed” ACSF functions, but not to “traffic jam assist” ACSF, since limited to lower speed
• Thus our proposal is to open ACSF “traffic jam assist” functions on all type of roads, including those without constructional separation, this on the following basis:
– Low speed traffic jam assist ACSF (below [40km/h]; without lane change capabilities) is not as safety critical as cruise speed ACSF (90km/h and lane change capabilities), thus the condition for constructional separation looks over-specified (unnecessary) for traffic jam assist
– The “use rate” of the function would be highly increased if permitted on all roads, which would relief truck drivers from very tiring and boring tasks, leading to unnecessary tiredness, which is indirectly positive for safety
– In a traffic jam, constructional separation may be hidden behind 1 or 2 lanes of stopped vehicles (if the truck is on the right lane, and the two other lanes on the left are crowded with vehicles), making the detection of the separation by the system erratic (which would lead to ACSF unwanted disabling)
Traffic Jam Assist on all roads
• A way forward could be to define different types of ACSF, with different requirement levels, for example:
ACSF type Requirements
ACSF < 10km/h same requirements as today
ACSF < [40km/h]ACSF with LK only
“traffic jam assist” allowed on all type of roads (even without constructional separation) provided lane change is disabled when used in traffic jam
ACSF < [130km/h]
ACSF with LK only ACSF with LK and Lane
Change
the system shall have some means to detect “highway conditions”. When the system detects the vehicle is on highway, the system is enabled.
Evaluation of ACSF duringperiodic technical inspection
Informal group ACSF #2
2015-06-16 / 17 in Tokyo
Informal Document ACSF-02-09
ITS-AD guidance to GRRFExtracts from ITS-AD_03-04-rev1
3. Possible discussion items on Automated Driving Technologies
The narrative definitions below have been taken from the SAE and can be used as a starting point to understand the level of assistance/automation:
[…] [Partial automation systems shall be so designed as to provide a continuous integrity check, recording any faults, failures, implausible messages etc., and shall record such events in a non-volatile memory. These data shall be accessible for the purposes of roadworthiness and maintenance inspection through a standardised scan tool. […]
3-2. Others
Discussion concerning electronic security, cybersecurity, roadworthiness inspection provisions (OBD), EDR, etc. could also be made in the IG-AD but should not preclude consideration by the appropriate GRs.
4. Guidance to GRRF (provisional draft)
Possible points to note
5) Adequate safety measure provision should be considered so as not to inhibit current development of such systems. These shall include, but not be limited to, HMI, system integrity monitoring, status recording.
OICA input:• The guidance opens up a number of items for possible discussions.• Recorded data are directly connected to roadworthiness and maintenance inspection.• OBD is only mentioned in the context of roadworthiness inspection provisions (PTI)
ACSF IG Terms of Reference
The terms of reference were endorsed at 79th session of GRRF
Extracts from GRRF-79 report (same as in ACSF-01-02)
2.The informal group shall address the following issues:
a) Review the current speed limitation (10 km/h) with the purpose of permitting ACSF functionality during [urban] and [interurban] journeys.
b) Define requirements for communicating to the driver a malfunction of ACSF.
c) Define requirements to enable the evaluation of ACSF during periodic technical inspection.
The focus of the ACSF group is on periodic technical inspection, which is quite aligned with the ITS / AD guidance
OICA input:1. Based on the TOR of the ACSF group, OICA has prepared proposals to fulfil the need for
ACSF evaluation during PTI.2. OICA proposal also covers the “real time” driver information of detected malfunctions.3. OICA proposal does not consider Maintenance aspects, since not in the TOR of the
group, and covered in other regulations (RMI)4. Yet OICA is open to discussions on other items mentioned in the ITS/AD guidance, for
further considerations in other relevant UNECE groups
Clarifications
OBD does not mean an “electronic interface to diagnose a vehicle with a scan tool”…
Below is reminded the definition of OBD in GTR 5:
3.12. "On-board diagnostic system (OBD)" means a system on board of a vehicle or engine which has the capability of detecting malfunctions, and, if applicable, of indicating their occurrence by means of an alert system, of identifying the likely area of the malfunctions by means of information stored in computer memory, and/or communicating that information off-board.
Based on this definition, UN safety regs (R13, R79, R131…) already includes OBD requirements:
• Capability of detecting malfunctions:
• Failure detection
• Failure classification (e.g. brake failures classified acc. to their impact on performance)
• Warning signals are an alert system
• Coarse identification of the area of the malfunctions: at least one warning per system/function, e.g.
• Braking yellow and red, tractor and trailers, ESC…
• Steering, LDWS, AEBS, TPMS etc.
OICA summary:Current requirements in UN regulations ensures safety in fault condition, based on failure detection and warning indication to driver in real time.
OICA Proposals for PTI - 1
Existing “OBD” requirements in safety UN regs ensure safety in fault condition, based on failure detection and warning indication to driver in real time.
They can easily be used for an efficient PTI
Up-to-date AEBS & LDWS regulations are suggested as a solid base for a starting point
CEL Annex guarantees performance of the steering system under fault and non fault conditions of the complex electronic system, including ACSF:
This is done via e.g. a description and a verification of the safety concept.
Extract: “In case of a failure, the driver shall be warned … by warning signal or message display”
In addition to the above, OICA is ready to consider the following extra-requirements, should they be judged necessary and justified by the CPs:
Request an electronic interface to confirm the “operational status of the ACSF system”(e.g. by reading the warning signal status or a system “roadworthiness” status…)
Such a requirement would permit the implementation of an efficient “fitment test” at PTI, to determine if the original system fitted from factory is still fitted to the vehicle and operational
This OICA proposal…1. Ensures safety and driver information 24/7 and fulfils the need for an efficient PTI2. Is perfectly matching with… New EU directive on PTI (applicable 2018)
German PTI regulation (deployment on-going by FSD)3. Can be implemented within a reasonable time frame compatible with industry targets
OICA Proposals for PTI - 2
Cornerstones:
The interface must be “read only”, to avoid safety and security issues (system damage, over-writing of calibration data, hacking…), leading to liability potential disputes:
Vehicle manufacturers cannot validate all PTI electronic tools tool vendors must take the responsibility of the above mentioned risks.
The only way is to keep safety OBD simple, to be safe and efficient
There must be no further design requirement on:
The connector type of the interface, other than those already included in the EURO VI / Euro6 OBD emission regulation.
Furthermore, the connector must remain open to other standards used e.g. in Japan.
The communication protocol of the interface other than those already applicable for diagnostic in ISO or SAE standards, e.g. SAE J1587, SAE J1939, ISO27145 etc.
The new PTI requirements on the electronic interface should apply only to new ACSF systems (above 10km/h), in order to avoid burdening existing / simpler technologies like e.g. LKAS or electronically controlled auxiliary axles, which are anyway our of the scope of this group
No Diagnostic Trouble Codes (DTC) list should be defined or required for PTI. DTCs are for repair, only the effect of DTC on performance is important for safety and roadworthiness of the vehicle.
• These are the conditions to keep PTI requirements compatible with industry and CPs targets on lead-time for ACSF implementation, while still ensuring safety and efficient PTI.
• They avoid adding a burden on systems not related to new ACSF functions
Proposal for draft amendment
5.5.2. It must shall be possible to verify in a simple way the correct operational status of those Complex Electronic Systems, which have control over steering. If special information is needed, this shall be made freely available. It must shall be possible to verify the correct operational status of those Electronic Systems by a visible observation of the failure warning signal status, following a "power-ON" and any bulb check.
In the case of the failure warning signal being in a common space, the common space must be observed to be functional prior to the failure warning signal status check.
[ In the case on an ACSF system able to operate at higher speed than 10km/h, it shall be possible to confirm the correct operational status given by the failure warning signal via the use on an electronic communication interface (i.e. the correct operational status shall be readable by an external electronic device). ]
5.5.2.1. At the time of Type Approval the means implemented to protect against simple unauthorized modification to the operation of the verification means chosen by the manufacturer (e.g. warning signal) shall be confidentially outlined.
Alternatively this protection requirement is fulfilled when a secondary means of checking the correct operational status is available, e.g. by using an electronic communication interface.
Base = AEBS textBold = changes to AEBS text
www.bmvi.de
Identification of regulatory needs for ACSF
Oliver Kloeckner
16-17th June 2015 2nd meeting of the IG ASCF
Tokyo – Jasic Office
Informal Document ACSF-02-10
Basics
Prerequisites for the following considerations are that
• in principle the driver is obliged to surveil the ACSF continuously and must be able to take over manual control at any time
• driving with ACSF shall be at least as safe as steering manually
Basics
• ACSF means continuous lateral control by automatic steering
• Automatic lateral control is needed for:
• following the path of a vehicle in front of the EGO-vehicle
• keeping the vehicle in a lane with orientiation by means of road markings
• lane changing manoeuvres
Basics
• The task is to derive general regulatory needs and performance requirements for ACSF
• Performance requirements have to cover:
a) „Normal“ driving with ACSF (carrying out the manouvres that ACSF was designed for)
b) The transition from manual steering to driving with ACSF and (even more important) the transition back to the driver when ACFS‘s system functionality boundary will be reached
c) Driver is or is going to be inattentive
d) Unexpected and critical events during driving with ACSF
HMI - Human capabilites to surveil
Establishing requirements we have to take into account:
• The capabilities of a human being degrade if workload is too high or too low
• The human being is not good in only surveilling a system
• Well working ACSF (possibly used togehter with ACC) may encourage the driver to do other tasks than controlling the vehicel or surveilling the ACSF
capabilities / vigilance / awareness
workload
over-load
under-load
Sources: see paper “HMI concept of ACSF - background knowledge from research”
HMI - Interaction between ACSF and driver
Design of ACSF has an influence on the driving task and thus on the capabilities of the driver (on vigilance, situation awareness, reaction times…)
Capabilities of the driver have to be taken into account when designing an ACSF and when establishing requirements for ACSF(ensuring mode awareness, signalling of system boundaries, design of warnings,…)
ACSF driver
HMI - Indication of ACSF system status
Needed: Clear and unambiguous (redundand) signalisation to the driver in whichstatus the ACSF is in order to achieve mode awareness at
• If the ACSF has detected that the conditions for safe automatically commanded steering are fulfilled it indicates to the driver that an activation is possible
• Driver activates if he likes to use the ACSF
• Deactivation or switch off is possible for the driver at any time
System off
System on and cannotbe activated
System on and can beactivated
System activated andworking
System requeststransition but still active
System on and cannotbe activated
Strategy to minimise risk if driver does not takeover manual steering again
Driver takes over steering
HMI - Reaction times of the driver Basics
1Sources:• see AEBS/LDW-06-08• see AEBS/LDW-07-05• further sources in paper “HMI concept of ACSF - background knowledge2depends considerably on traffic situation, time since automated function started, driver capabilites, driver age,…3reseach to determine human reaction while driving with automated driving tasks has just started!
driver status Reaction time1
Full situation awareness About 0.7 s to 1.4 s
Manual driving but distracted About 1 s to 2 s
Only monitoring About 5 s(study, simulator)
Other task than driving becomes primarytask
Can be much longer2, 3
www.bmvi.de
General Requirements
General objectives:• Safe path follwing• Safe lane keeping• Safe lane change
• ACSF should not cause safety critical situations• Steering manoeuvres must be clear to other road users
Needed:• Limit lateral acceleration to [3] m/s2 in order to avoid high dynamic
manoeuvres• Minimum distance (e. g. in terms of minimum time gap) to other road
users in all driving situations at any time (not only laterally but also longitudinally, Interaction with the brake seems necessary)
• direction indicators must be activated automatically in case of ACSF lane change manoeurvres
a) Normal driving with activated ACSF
b) Transition from ACSF to manual steering
Reaching the ACSF system boundaries will result in a transfer of the steering task back to the driver e. g. in the situations (depending on ACSF system functionality)
• Specified speed boundaries are reached• End of congestion / lead vehicle gone• Sensor‘s capabilites reduced (e.g. snow)• Constucion site / lane too narrow / end of lane• Road marking with degraded quality• Road curvature too tight• …
It must be ensured that safe transition to manual steering is possible at any time Needed: If system will not be able to cope with oncoming circumstances:
• Early request to the driver that that transition from ACSF to manual steering is necessary and that the driver must take over control of the steering task
• Request using at least 2 sensory channels, escalating warning
b) Transition from ACSF to manual steering
If the driver does not take over the steering after a transition request was given by the system
• ACSF must comprise some strategy to reach a status with as less risk as possible in the given traffic situation
Needed e. g.:
• Further lane keeping for a certain time• Enlarging gap to other road users• Slowing down to standstill• Switching hazard lights on• If lane change is part of ACSF‘s system functionality: lane change to edge of the
road• …• (or a combination of the above mentioned actions)
c) Driver is going to be inattentive
• ACSF shall ensure the drivers attantiveness
• Means to detect drivers attentiveness(e.g. Driver absence detection, Hands off detection, Torque sensing)
• Means to maintain or regain situation awareness• If inattentiveness remains: start transition procedure
• ACSF shall have minimum precausionary measures to avoid to misuse
• Warning and Transition request if driver seatbelt is unfastened• Warning and Transition request if driver seat is left
d) Unexpected critical events
The ACSF shall be able to cope with unexpected and sudden critical events (time is too short for safe transition procedure)
e. g. in the cases:• Obstacle in lane• Accident is happening just ahead• Pedestrian steps on the road• Emergency braking of the vehicle in front• Cutting-in vehicle is going to cause side collision• Lane change manouvre of EGO-vehicle coincidents with lane change of another
vehicle targeting the same lane at the same positionNeeded:
• Automatic emergency braking• Abortion of manoeuvres (e. g. in lane change)• Evasive steering (on manufacturer‘s choice)
www.bmvi.de
Possible Tests
Test of functionality (advanced lane keeping)
Check if vehicle stays on path under normaloperating conditions (ACSF on and activated)
Test FU1• at least 5 min driving behind a lead vehicle (ACSF on and activated)
(time gap: 2 s to 3 s for manual speed adjustment, otherwise time gap is selected by the vehicle itself)
• lead vehicle shall drive within the lane markings with ay ≤ 1 m/s²• on a track with road markings of good visibility at each side of the lane• at various speeds up to vsmax and down to vsmin• with various curvatures
Test is passed: vehicle does not cross any lane marking
Proposal - Test of functionality (advanced lane keeping)
Check if vehicle stays on path under normaloperating conditions (ACSF on and activated)
Test FU2• at least 5 min driving without a lead vehicle (ACSF on and activated)• with ay ≤ 1 m/s²• on a track with road markings of good visibility at each side of the lane• at various speeds up to vsmax• with various curvatures
Test is passed: vehicle does not cross any lane marking
Test of transition request at system boundaries and minimum risk manoeuvre
tight curve: ay beyond system boundaries
with or without lead vehicle
Test TR1• driving at least 1 min behind a lead vehicle (ACSF on and activated)• lead vehicle shall drive within the lane markings• on a track with road markings of good visibility at each side of the lane• at a speed of 10 km/h below vsmax• after a straight section of at least 200 m the vehicle shall approach a curve of
more than 90° that would demand an ay of more than [3] m/s²• test driver shall not take over manual steering control again
Test is passed:• transition request was given at least when the lateral acceleration exceeds [3]
m/s²• the minimum risk maneuver as specified by the manufacturer was initiated• vehicle does not cross any lane marking before the minimum risk maneuvre
was initiated.
Test TR2• driving at least 1 min without a lead vehicle (ACSF on and activated)• on a track with road markings of good visibility at each side of the lane• at a speed of 10 km/h below vsmax• after a straight section of at least 200 m the vehicle shall approach a curve
of more than 90° that would demand an ay of more than 3 m/s2 • test driver shall not take over manual steering control again
Test is passed:• transition request was given at least when the lateral acceleration exceeds [3]
m/s²• the minimum risk maneuver as specified by the manufacturer was initiated• vehicle does not cross any lane marking before the minimum risk maneuvre
was initiated.
Test of transition request at system boundaries and minimum risk manoeuvre
lane markings quality beyond system boundaries
with or without lead vehicle
Test TR3• driving at least 1 min behind a
lead vehicle (ACSF on and activated)• lead vehicle shall drive within the lane markings• on a track with road markings of good visibility at each side of the lane• at a speed of 10 km/h below vsmax
• after a straight section of at least 200 m the vehicle shall approach a section with a length of 200 m with only one lane marking at the driver‘s side
• test driver shall not take over manual steering control again
Test is passed:• vehicle does not cross any lane marking and• the transition request was given [0 s] before the vehicle would have entered the
section with missing lane marking and• the minimum risk maneuver as specified by the manufacturer was initiated
Test TR4• driving at least 1 min without a lead vehicle
(ACSF on and activated)• on a track with road markings of good visibility at each side of the lane• at a speed of 10 km/h below vsmax• after a straight section of at least 200 m the vehicle shall approach a section
with a length of 200 m with only one marking on the driver‘s side • test driver shall not take over manual steering control again
Test is passed:• vehicle does not cross any lane marking and• the transition request was given [5 s] before the vehicle would have entered the
section with missing lane marking and• the minimum risk manoeuvre as specified by the manufacturer was initiated
Emergency Tests
Test Em1
Basic automatic emergency braking capability if leadvehicle suddenly decelerates sharply (ACSF on and activated)
• Test speed of both vehicles 10 km/h below vsmax• Initial time gap 3 s• Lead vehicle deceleration 6 m/s², mean jerk 6 m/s³
Test is passed: No collision
6 m/s2
3 s
Emergency TestsTest EM2
Basic automatic emergency braking capability if pedestrianstands statically in the lane (ACSF on and activated
• Test speed 10 km/h below vsmax• Pedestrian centered in the lane
Test is passed: No collision
Pedestrian dummy
Emergency TestsTest EM3
Abortion capability if second lane is suddenly occupied by another vehicle (ACSF on and activated)
• ego vehicle drives between two other vehicles in same lane, all at [30 km/h] below vsmax or at vsmin, whatever is higher (perhaps test at lower speeds for safe testing necessary)
• time gap to two other vehicles 3 s• during lane change third vehicle overtakes with 50 km/h above vsmax• TTC of third vehicle at point in time when vehicle under test is crossing lane: [2 s]
Test is passed: Abortion of overtaking manoeuvre and lane keeping continues
www.bmvi.de
Thank you for your attention!
Federal Ministry of Transport and Digital Infrastructure
Robert-Schuman-Platz 1D-53175 Bonn
Submitted by the expert from Germany
Basic sketch of dependencies to identify regulatory needs
driver is attentive driver is not attentive
"normal" driving with ACSF
early request for transition
strategy to minimise risk
activation
successful
detection of inattentiveness system boundary will be
reached
warning means to regain
attentiveness
successful
remains inattentive
steering manually
deactivation
not successful
inattentiveness is not detected
properly
at any time there should be a fall back for sudden critical events (although sometimes a good and attentive driver would be able to master the situation):
system has to comprise a strategy to cope with sudden critical situations with the aim to avoid collisions (e. g. automatic emergency braking et al.)
Informal Document ACSF-02-11
Necessity of Event Data Recorders (EDRs) and
On-board Diagnostic (OBD) Systems for ACSF
1. Background
In ACSF, the system affects the lateral movement of the vehicle at a greater extent than
existing advanced driver assistance systems. The system operates the vehicle in part,
instead of the driver, under driver’s monitoring, which means that higher weight of the
vehicle control would be transferred from the driver to the system. (It does not
immediately mean that the responsibility of the driving is transferred to the systems.)
The electronically-controlled systems for ACSF are so sophisticated and so complicated
than existing advanced driver assistance systems that car users cannot find the
malfunctions of the system which may cause safety risk. This new relationships between
the driver and the system may cause new types of risks, as follows.
In the case of a significant malfunction, it may be more technologically difficult to
control the vehicle’s behavior during lateral control than during longitudinal control.
Safety risk may be caused during the transition from the system to the driver.
It would be difficult for car users to find and repair the malfunctions because the
system is combined with several functions such as lateral control longitudinal
control.
2. Proposal
EDR and OBD shall be necessary for ACSF in order to ensure the safety in use.
(1) EDR:
By recording the vehicle system’s performance as well as the driver’s operation before
and after the relevant event, it will enable the verification, whether the system was
functioning properly.
=> If it is found that the system caused the safety risk, the appropriate remedial
measures should be taken without delay..
*Note: With regard to EDR, applicable rules on privacy information protection shall be
respected.
2) OBD:
By monitoring the system, it will detect malfunctions and inform the driver of it. It will
also retain the log of the malfunctions.
=> In case that a relevant malfunction occurs in ACSF, the driver shall disable the ACSF
so that the system will not lead safety risk. It is also important to prepare the
environments where the car users can get their ACSF repaired at any repair stations
(not only at car dealers but also at the other repair shops).
Informal Document ACSF-02-12
Submitted by the expert from Germany
Proposal for amendments to Regulation No. 79 to include ACSF > 10 km/h
The modifications to the Regulation are marked in blue bold and strikethrough characters.
==================================================================================
Amend paragraph 2.3.4.1. to read:
2.3.4.1. "Automatically commanded steering function" (ACSF) means the function within a
complex electronic control system where actuation of the steering system can result
from automatic evaluation of signals initiated on-board the vehicle, possibly in
conjunction with passive infrastructure features, to generate continuous control
action in order to assist the driver in following a particular path, in low speed
manoeuvring or parking operations.
2.3.4.1.1. Category 1 ACSF means, a function that operates at a speed no greater than 10
km/h to assist the driver, on demand, in low speed manoeuvring or parking
operations.
2.3.4.1.2. Category 2 ACSF means, a function that operates at a speed no greater than [130 km/h] and which can perform a single manoeuver (e.g. lane change) when commanded by the driver. (B): single manoeuver? Is this for the overtaking of only one car or one manoeuver
in which several cars are passed? is it just about a lane change? Does this possibility
has to be activated before each manoeuver or can this be activated on a long
stretch of road?
2.3.4.1.3. Category 3 ACSF means, a function that operates at a speed no greater than [130
km/h] and which can indicate the possibility of a single manoeuvre (e.g. lane
change) but performs that function only following a command by the driver.
(B): What's exactly the difference with category 2 ACSF?
2.3.4.1.3. Category 4 ACSF means, a function that operates at a speed no greater than [130
km/h], which is commanded by the driver and which can continuously determine
the possibility of a manoeuvre (e.g. lane change) and complete these manoeuvers
for extended periods without further driver command.
Insert new paragraph 2.4.8. to read
2.4.8. For Automatically commanded steering functions
2.4.8.1 “Motorway” means, a road section, dedicated exclusively to motor vehicles, having at least two
traffic lanes for each direction of travel and having a physical separation of traffic moving in
opposite directions.
(B): Better to use the definition of a motorway as stated in the convention on road traffic
of 1968, cf. art. 1, j)
Informal Document ACSF-02-03
Informal Document ACSF-02-13
Submitted by the expert from Belgium (in red)
2.4.8.2 "Lane" means one of the longitudinal strips into which a roadway is divided.
(B): carriageway is the correct term, cf. art. 1, e) convention on road traffic of 1968
2.4.8.3 "Visible Lane markings" means delineators intentionally placed on the borderline of the
lane that are directly visible by the driver while driving (e.g. not covered by snow, etc.).
(B): broken lines, continuous lines or other appropriate means indicating the traffic lanes
(cf. convention on road signs and signals, annex 2, chapter II, B.
2.4.8.4 "Lead vehicle" means a vehicle driving in front of the vehicle equipped with ACSF.
2.4.8.5 "Lane change manoeuvre" means a manoeuvre in which the vehicle changes from its initial
travel lane to an adjacent lane (B): traffic lane
2.4.8.6 "Specified maximum speed Vsmax " means the maximum speed up to which an ACSF is
designed to work.
2.4.8.7 "Specified minimum speed Vsmin" means the minimum speed up to which an ACSF is
designed to work.
2.4.8.8 "Transition request" means a request to the driver that the driver has to take over manual
control of the steering task again.
2.4.8.9 "Transition procedure" means the sequence of providing a transition request by the ACSF,
taking over manual steering control by the driver and deactivation of the ACSF since
manual control was detected by the ACSF.
2.4.8.10 "Conditions for safe operation" mean all circumstances like traffic situation, road
category, quality of lane markings, vehicle speed, curvature of the road, lighting, sensor
capacities etc. specified by the vehicle manufacturer that have to be fulfilled when an ACSF
shall be able to be activated by a driver.
2.4.8.11 "System boundaries" mean all circumstances from which on the conditions for safe
operation are not fulfilled anymore, that cannot be dealt with by an activated ACSF
anymore and thus request a take-over of manual steering control by the driver.
2.4.8.12 "ACSF status" means any distinct operational mode of the ACSF like "switched off"
"switched on", "available to be activated", "activated" etc.
2.4.8.13 "Attention recognition system" means a device to detect if the driver is vigilant, is
attentive, is aware of the traffic situation
2.4.8.14 "Minimum risk manoeuvre" means a strategy for the [longitudinal and] lateral control of
the vehicle to reach a status with as little risk as possible in the given traffic situation when
the driver is detected by the ACSF not to be available.
Amend paragraph 5.1.6.1. to read:
5.1.6.1. Whenever the an Automatically Commanded Steering function becomes operational, this
shall be indicated to the driver. and the control action shall be automatically disabled if the
vehicle speed exceeds the set limit of 10 km/h by more than 20 per cent or the signals to be
evaluated are no longer being received. Any termination of control shall produce a short but
distinctive driver warning in accordance with the requirements of paragraph 5.4.3. by a
visual signal and either an acoustic signal or by imposing a tactile warning signal on the
steering control.
Insert new paragraph 5.4.3. Renumber paragraph 5.4.3. as 5.4.4.
5.4.3. Special Warning Provisions for Automatically Commanded Steering Functions
5.4.3.1 Any termination of control shall produce a distinctive driver warning by a [yellow] visual
signal and either an acoustic signal or by imposing a haptic warning signal. This warning
shall be provided before the system (function) becomes in-operational, if the termination
is not intended by the driver. (B): Add - "This warning shall remain operational until the
driver has resumed the control."
5.4.3.2. Any sudden termination of control caused by a failure of the system physical or functional
failure shall produce immediately a short but distinctive driver warning by a [red] visual
signal and either an acoustic signal that shall remain operational until the driver has
resumed control.
Insert new paragraph 5.6
5.6 Special Provisions for Automatically Commanded Steering Functions
5.6.1. Special Provisions for ACSF of Category 4
5.6.1.1. General
5.6.1.1.1. The vehicle shall be equipped with a means for the driver to activate or deactivate
the ACSF at any time.
5.6.1.1.2. The ACSF shall be activatable only if the conditions for safe operation of the ACSF
are fulfilled [all associated functions – brakes, accelerator, steering,
camera/radar/lidar etc. are working proper).
5.6.1.1.3. The ACSF shall only activate by deliberate action of the driver.
5.6.1.1.4. The ACSF shall be able to detect if the driver controls the steering function
manually. If the ACSF detects, that the driver is steering manually , ACSF shall be
deactivated.
5.6.1.1.5. The ACSF may be operational up to a vehicle lateral acceleration of [3] m/s2.
5.6.1.1.6. The ACSF shall comprise an attention recognition system that is active whenever
the ACSF is active.
5.6.1.2. Operation of ACSF
5.6.1.2.1. Any lane change manoeuvre shall be initiated only if:
- the vehicle is travelling on motorway as defined in paragraph 2.4.8. and
- any traffic that can affect the safe manoeuvre shall be identified by equipment
installed on the vehicle and
- the vehicle equipment can analyze speed and distance of the identified traffic to
ensure a safe manoeuvre(e.g. does not cause a deviation to the flow or direction of
other traffic).
5.6.1.2.2. If a lane change manoeuvre is carried out, the correspondent direction indicator
lamps shall be automatically activated.
5.6.1.2.3. If the ACSF is not overridden by the driver it shall not terminate the lane change
until the manoeuvre is safely completed, except for the ACSF detects an imminent
critical situation. (B): what does this mean exactly? Difficult to understand, I think
the word "not" should be deleted in the first phrase.
5.6.1.2.4. The activated ACSF shall at any time control the lateral movements of the vehicle in
such a way that the vehicle does not induce any safety critical situations and that
the movements of the vehicle are clear to other road users.
5.6.1.2.5. The activated ACSF shall at any time ensure a safe lateral distance to other road
users. The vehicle manufacturer shall provide documentation about how such a
safe distance is achieved to the technical service.
5.6.1.2.6. The ACSF shall be designed such that safe transition to manual steering is possible
at any time.
5.6.1.2.7. If the activated ACSF detects that due to a sudden unexpected event the vehicle is
in imminent danger to collide with another road user and that the time for a safe
transition procedure is too short, an emergency manoeuvre shall be carried out.
The vehicle manufacturer shall provide information to the technical service about
the safety strategy depending on different circumstances forming a sudden critical
event and the foreseen emergency manoeuvres.
(B): Is it up to each manufacturer to define the emergency manoeuvre? Shouldn't
this not be defined within this regulation?
5.6.1.2.8. If the attention recognition system detects that the driver is inattentive, it shall give
a warning to restore attentiveness again. The manufacturer shall provide
information to the technical service how the attention recognition systems detects
inattentiveness of the driver. (B): Add - "This warning shall be given until the
moment that the driver is attentive again."
5.6.1.3. System boundaries
5.6.1.3.1. The vehicle manufacturer shall provide the values for Vsmax and Vsmin to the technical
service.
5.6.1.3.2. The vehicle manufacturer shall provide an information to the technical service
under which conditions an ACSF can be activated, i. e. when the conditions for safe
operation of the ACSF are fulfilled.
5.6.1.3.3. The vehicle manufacturer shall provide information to the technical service about
system boundaries at which the activated ACSF must give a transition request.
5.6.1.4. Indication of ACSF status
5.6.1.4.1. The ACSF shall at any time give a noticeable and distinctive signalization to the
driver about the ACSF status. This signalization shall be at least a visual signal. Any
change in system status shall be indicated by an optical and either an acoustic or
haptic signal.
5.6.1.5. Transition request
5.6.1.5.1. If ACSF detects that its system boundaries are reached or will be reached shortly it
shall provide a transition request.
5.6.1.5.2. The timing of the transition request shall be such that sufficient time is provided for
a safe transition of the steering task from automatically commanded steering to
manual steering.
5.6.1.5.3. The vehicle manufacturer shall provide specific values for time intervals to the
technical service, which are foreseen for safe transition under different
circumstances.
(B): Is it up to manufacturers to determine the transition period, shouldn't this be
part of the regulation, for example - when the vehicle is driving at a speed of xx
km/h the transition period is xxx seconds, etc., etc. This should be standardized.
5.6.1.5.4. If the speed of the vehicle with activated ACSF exceeds vsmax a transition request
shall be given.
5.6.1.5.5. If the vehicle reaches a lateral acceleration of more than [3] m/s2 a transition
request shall be given.
5.6.1.5.6. If an attention recognition system detects the driver to be inattentive although a
warning to restore attentiveness was provided to the driver a transition request
shall be given.
(B): meaning that if the driver is inattentive he will be have to take over the
control? Is this logic? Wouldn't it be better to keep on warning the driver up until
the moment that he's attentive again rather than give him a transition request?
5.6.1.5.7. The ACSF shall provide a transition request if the driver's seatbelt is unfastened
and/or if the driver's seat is left by the driver.
5.6.1.5.8. The transition request shall be provided by a [yellow] visual signal and either an
acoustic signal or by imposing a haptic warning signal.
5.6.1.5.9. If the driver does not take over manual control immediately the transition request
shall be escalating with time in terms of enlarging the intensity of the warning
and/or in terms of adding and/or changing the warning means.
5.6.1.5. Minimum Risk Manoeuvre
5.6.1.5.1. If the ACSF detects that after a transition request the driver does not take over
manual control of the steering again the vehicle shall carry out a minimum risk
manoeuvre. The vehicle manufacturer shall provide information to the technical
service about which kind of minimum risk manoeuvres are foreseen depending on
the given traffic situation and circumstances at its initiation. (B): see, 5.6.1.2.7.
5.6.2. Special Provisions for ACSF of Category 3
- to be developed based on the requirements for a Category 4 system-
..
5.6.3. Special Provisions for ACSF of Category 2
- to be developed based on the requirements for a Category 4 system-
..
5.6.4. Special Provisions for ACSF of Category 1
- to be developed based on the requirements for a Category 4 system-
.
Insert new Annex 7
Annex 7
Text requirements for automatically commanded steering functions
1. General Provisions
Vehicles fitted with ACSF shall fulfill the tests requirements of this annex according to the
corresponding category of ACSF specified in Table 1.
2. Test conditions
2.1. The test shall be performed on a flat, dry asphalt or concrete surface delivering good adhesion. The ambient temperature shall be between 0° C and 45° C.
2.2. Lane markings
The lane markings and the width of the lane used in the tests shall be those of one of the Contracting Parties, with the markings being in good condition and of a material conforming to the standard for visible lane markings of that Contracting Party. The lane marking layout used for the testing shall be recorded.
The test shall be performed under visibility conditions that allow safe driving at the required test speed.
2.3 Lead vehicle
The lead vehicle shall be a high volume series production passenger car of category M1 AA saloon or in the case of a soft target an object representative of such a vehicle in terms of its detection characteristics. A soft target is a target that will suffer minimum damage and cause minimum damage to the subject vehicle in the event of a collision.
2.4 Pedestrian soft target
A pedestrian soft target is an object representative of a human adult in terms of its detection characteristics that will suffer minimum damage and cause minimum damage to the subject vehicle in the event of a collision.
3. Tests
Table 1 specifies which tests have to be fulfilled by each ASCF category.
Test \ ACSF category 1 2 3 4
FU1 path following with lead X
FU2 lane keeping w/o lead X
TR1 tight curve with lead X
TR2 tight curve w/o lead X
TR3 no marking with lead X
TR4 no marking w/o lead X
EM1 AEB on braking rabbit X
EM2 AEB on static pedestrian X
EM3 abortion of lane change X
Table 1
3.1. Functionality Tests
3.1.1. Functionality Test 1 (FU1)
Drive the vehicle with activated ACSF at least 5 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with various
curvatures with road markings at each side of the lane at various speeds up to vsmax and
down to vsmin. The speed of the lead vehicle shall be selected such that the lateral
acceleration is not more than 1 m/s².
The requirements of the test are fulfilled if the vehicle does not cross any lane marking.
3.1.2. Functionality Test 2 (FU2)
Drive the vehicle with activated ACSF at least 5 min without a lead vehicle. The vehicle shall
drive on a track with various curvatures with road markings at each side of the lane at
various speeds up to vsmax and down to vsmin. The speed shall be selected such that the
lateral acceleration is not more than 1 m/s².
The requirements of the test are fulfilled if the vehicle does not cross any lane marking.
3.2. Transition Tests
3.2.1 Transition Test 1 (TR1)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with road markings
at each side of the lane at a speed of 10 km/h below vsmax. After a straight section of at
least 200 m the lead vehicle shall enter a curve of more than 90° that demands a lateral
vehicle acceleration of more than 3 m/s2. The test driver of the vehicle shall not take over
manual steering control again.
The requirements of the test are fulfilled if the transition request was given at least when
the lateral acceleration exeeds [3] m/s² and the minimum risk maneuvre as specified by
the manufacturer was initiated. The vehicle shall not cross any lane marking before the
minimum risk maneuvre was initiated.
3.2.2. Transition Test 2 (TR2)
Drive the vehicle with activated ACSF at least 1 min without a lead vehicle. The vehicle shall
drive on a track with road markings at each side of the lane at at a speed of 10 km/h below
vsmax. After a straight section of at least 200 m the vehicle shall approach a curve of more
than 90° that would demand a lateral vehicle acceleration of more than 3 m/s2. The test
driver of the vehicle shall not take over manual steering control again.
The requirements of the test are fulfilled if the transition request was given at least when
the lateral acceleration exeeds [3] m/s² and the minimum risk maneuvre as specified by
the manufacturer was initiated. The vehicle shall not cross any lane marking before the
minimum risk maneuvre was initiated.
3.2.3. Transition Test 3 (TR3)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the ACSF, the vehicle shall drive between [2 s] and [3 s] behind the lead
vehicle. The lead vehicle shall drive within the lane markings on a track with road markings
at each side of the lane at a speed of 10 km/h below vsmax. After a straight section of at
least 200 m the lead vehicle shall enter a section with a length of 200 m with only one lane
marking at the driver‘s side. The test driver of the vehicle shall not take over manual
steering control again.
The requirements of the test are fulfilled if the vehicle does not cross any lane marking and
if the transition request was given [0 s] before the vehicle would have entered the section
with missing lane marking and the minimum risk maneuvre as specified by the
manufacturer was initiated.
3.2.4 Transition Test 4 (TR4)
Drive the vehicle with activated ACSF at least 1 min without a lead vehicle. The vehicle shall
drive on a track with road markings at each side of the lane at at a speed of 10 km/h below
vsmax. After a straight section of at least 200 m the vehicle shall approach a section with a
length of 200 m with only one lane marking at the driver‘s side. The test driver of the
vehicle shall not take over manual steering control again.
The requirements of the test are fulfilled if the vehicle does not cross any lane marking and
if the transition request was given [5 s] before the vehicle would have entered the section
with missing lane marking and the minimum risk maneuvre as specified by the
manufacturer was initiated.
3.3. Emergency Tests
3.3.1 Emergency Test 1 (EM1)
Drive the vehicle with activated ACSF at least 1 min behind a lead vehicle. If the time gap is
not selected by the system, the vehicle shall drive at a gap of 3 s behind the lead vehicle.
The lead vehicle shall drive within the lane markings on a track with road markings at each
side of the lane at a speed 10 km/h below vsmax. Then the lead vehicle decelerates with 6
m/s² and with a mean brake jerk of 6 m/s³ in the first second of braking.
The requirements of the test are fulfilled if the vehicle does not collide with the lead
vehicle.
3.3.2. Emergency Test 2 (EM2)
Drive the vehicle with activated ACSF at least 1 min on a track with road markings at each
side of the lane at a speed 10 km/h below vsmax. The vehicle shall approach a stationary
pedestrian soft target being placed in the center of the lane.
The requirements of the test are fulfilled if the vehicle does not collide with the pedestrian
soft target.
3.3.3. Emergency Test 3 (EM3)
Drive the vehicle with activated ACSF at least 1 min on a track with road markings at each
side of the lane at a speed of [30 km/h] below vsmax or at vsmin, whatever is higher, between
two other vehicles. If the time gap is not selected by the system, the vehicle shall drive at a
gap of 3 s behind the lead vehicle. Induce by e.g. selecting a higher desired speed a lane
change manoeuvre. During the lane change a further overtaking vehicle with a speed of 50
km/h above vsmax shall approach from backwards on the adjacent lane that was intended to
be reached with the lane change. At the point in time when the vehicle is crossing the lane
marking the further overtaking vehicle shall be at a TTC of [2 s] behind the vehicle.
The requirements of the test are fulfilled if the vehicle does not collide with the further
overtaking vehicle or any other vehicle and aborts the lane change manoeuvre.