Design and verification of driver interfaces for adaptive cruise control

Design and verification of driver interfaces for adaptive cruise control
Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
www.springerlink.com/content/1738-494x(Print)/1976-3824(Online)
DOI 10.1007/s12206-015-0536-9
Design and verification of driver interfaces for adaptive cruise control systems†
Sang Hun Lee* and Dae Ryong Ahn
Graduate School of Automotive Engineering, Kookmin University, Seoul, 136-702, Korea
(Manuscript Received November 12, 2014; Revised December 28, 2014; Accepted January 7, 2015)
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Abstract
In a highly automated system, mode confusion is a significant human error factor that contributes to accidents. To suppress mode confusion in adaptive cruise control (ACC) systems of vehicles, we developed a new driver interface based on a formal approach to analyze
and verify human-automation interaction. To enhance the driver’s mode awareness, we developed a new ACC interface that eliminates
inconsistent mode transitions by reconfiguring the modes. Then, a human-in-loop experiment was conducted in a simulated environment
where a driving simulator was used to evaluate the state and mode awareness of drivers with the old and new interfaces. The experimental results showed that the proposed interface model, which was verified a formal method, was very effective in reducing mode confusion
compared with the traditional interface model.
Keywords: Adaptive cruise control; Formal verification; Intelligent vehicle; Mode confusion; User interface
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Introduction
Recently, intelligent vehicles have been equipped with
adaptive cruise control (ACC) systems to enhance the safety
of drivers and passengers. ACC systems enhance conventional
cruise control systems and they are capable of controlling the
speed of a vehicle, which is set by the driver. This allows the
host vehicle to follow a forward vehicle at an appropriate distance by controlling the engine and/or power train, and potentially the brake. According to the test specifications of ISO
15622:2010, three operational modes are found in ACC systems, i.e., off, standby, and active, while the active mode also
has sub-modes such as speed control and following control [1].
Therefore, it is necessary to consider how to reduce the possibility of mode confusion during the design of driver interfaces
for ACC systems.
There has been a lot of research on mode confusion and its
resulting automation surprise of highly automated systems, in
particular, aircrafts [2-7]. Mode confusion in ACC systems
has been investigated by several researchers recently. Horiguchi et al. [8, 9] showed that if different modes exhibit a similar
response, users have difficulty distinguishing them. Thus, they
proposed a new method that uses mode vectors to estimate the
possibility of mode confusion. Furukawa et al. [10] conducted
an experimental study of mode awareness using a dual-mode
ACC system with high-speed and low-speed range modes.
*
Corresponding author. Tel.: +82 2 910 4835, Fax.: +82 2 910 4718
E-mail address: shlee@kookmin.ac.kr
†
Recommended by Associate Editor Ki-Hoon Shin
© KSME & Springer 2015
They determined the information that could be effective for
supporting the mode awareness of driver in complex situations
if some direct information on the system states is concealed.
Heymann and Degani [11] described a hierarchy of automated
driving aids and their functionalities, beginning with standard
cruise control (CC), followed by ACC with the option of full
speed range functionality to allow stopping and starting, before adding automatic lane centering, and the possible future
incorporation of navigation functions. Ahn et al. [12] studied
possible mode confusions in a simulated environment when
vehicles are equipped with an adaptive cruise control system.
A set of human-in-loop experiment was conducted to observe
possible mode confusions and redesign the user interface to
reduce. Lee et al. [13] also developed a new driver interface
for the ACC system to suppress mode confusion, based on a
formal method.
In order to reduce mode confusion, it is necessary to provide the driver with a transparent display of the automation
state, as well as correct and succinct information via the user
interface [14, 15]. To satisfy these requirements, in this study,
we analyzed a conventional driver interface for ACC systems
in a formal method and we examined the occurrence of mode
confusion based on human-in-the-loop experiments. In addition, a new driver interface was developed, and it was verified
for its improvement in terms of mode confusion via formal
verification and human-in-the-loop experiments. The design
and verification process ensured that the user’s mental model
was correct and succinct, while the information display was
sufficiently clear to prevent mode confusion.
2452
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
Fig. 1. Machine model and interface model of the ACC system.
This paper is organized as follows. Sec. 1 introduces the
background and objectives of this study, including a survey of
previous research related to mode confusion and driver interfaces in ACC systems. In Sec. 2, we present the results of a
test of a conventional driver interface for an ACC system,
where we used a formal analysis method to determine whether
its cognitive model of users might cause mode confusion. We
also analyzed whether the interface model was consistent with
the standardized operational modes. In Sec. 3, based on the
results, we developed a new driver interface as an alternative
that is more robust to mode confusion, where we followed the
rules for clear user interface design. In Sec. 4, we describe the
results of driver mode awareness tests in the existing and new
driver interfaces based on human-in-the-loop experiments in a
simulated environment. In Sec. 5, we summarize the experimental results and discuss them in detail. In Sec. 6, we conclude the study and suggest future research.
2. Machine and user models for the ACC system
2.1 Conventional machine model for the ACC system
As illustrated in Fig. 1, the machine model for the existing
ACC systems implemented in most current vehicles comprises
six states: off, armed, canceled, override, speed-control, and
gap-control. The user-triggered events include pushing and
releasing the brake and gas pedals, and pushing various buttons to activate, cancel, and resume the ACC, as well as increasing and decreasing the set speed and set distance. If the
driver turns on the ACC system by pushing the ACC button,
the ACC enters the armed state, which is a type of ready state
that waits for specific activations by the driver. When the set
button is pushed, the current speed is set as the set speed and
the state is switched to speed-control. A constant vehicle
speed is maintained using the set speed in the speed-control
state. However, if a vehicle is found in front within the set
distance, the state is changed automatically to gap-control,
where the host vehicle follows the target vehicle at a constant
distance. Of course, the state returns to the speed-control when
the target vehicle exceeds the gap-distance range. In the gapcontrol or speed-control states, the driver can take control of
the vehicle from the ACC by pushing the gas pedal. This is
called the override state. The state returns to speed-control as
soon as the gas pedal is released in the override state. The
canceled state is a standby state in ACC, which occurs when
the brake is applied or the cancel button is pushed in the gapcontrol or speed-control states. The resume button should be
pushed by the driver in order to recover the speed-control state
from the canceled state.
2.2 Conventional interface model for the ACC system
The interface model for the ACC system, which was suggested in the test specification of ISO 15622:2010 [1] and is
used most widely in existing automobiles, provides three
modes, i.e., active, standby, and off, where transitions occur
between modes via user-triggered events, as shown in Fig. 1.
A single mode in an interface model actually includes one or
multiple states in a machine model. In the existing interface
model, as shown in Fig. 1, the gap-control, speed-control, and
override states are located in the active mode, the canceled
and armed states are in the standby mode, and the off state is
in the off mode.
2.3 Conventional user interface for the ACC system
In this study, a user interface was implemented based on the
interface of the ACC system used by the major automotive
manufacturers including Hyundai [16]. In Fig. 2, “ACC” and
“Set 70 Km/h” are parts that represent a mode in the ACC
system, while the figure in the center of the gauge cluster
represents the internal states of the active mode. As shown in
Table 1, “ACC” distinguishes the off mode from the other
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
Fig. 2. User interface of the ACC system implemented based on that of
the Hyundai Equus [13].
Table 1. User interface for each mode in the existing ACC system [13].
2453
should not be a problem because they make transitions into
states in the same mode. However, if they make transitions
into different modes, the users may be confused and surprised
by incompatible mode changes. These cases are indicated by
“X” in the final column: “Incompatible.” Incompatible transitions may occur in the standby and active modes. When the
resume button is pushed, the canceled (CAN) and armed
(ARM) states in the standby mode can make transitions into
the speed-control state in the active mode, or remain in the
armed state in the standby mode, thereby resulting in transitions into different modes. Because the armed and canceled
states in the same mode described above have the same display, a driver cannot distinguish them. Moreover, in the active
mode, the override state is incompatible with the gap- and
speed-control states because the brake pedal is not available in
the override state whereas it is available in the other states.
Therefore, the driver becomes confused and surprised by the
unexpected incompatible modes produced by the same user
input. Furthermore, although mode confusion does not occur
when the override state belongs to the active mode, it is not
acceptable for users if the override state is grouped into the
active mode because the ACC system loses control of the
vehicle in the override state.
3. Development of a new user interface for the ACC
system
modes. “Set__km/h” distinguishes the active and standby
modes; thus, if “Set__km/h” is turned off, this indicates the
standby mode, whereas, if “Set__km/h” is turned on, this indicates the active mode.
2.4 Conventional user interface for the ACC system
The state transitions triggered by the driver’s direct operations in the ACC system can be summarized by a state transition table, as shown in Table 2. The same user-triggered operation is applied to the states that belong to the same mode,
but they may make transitions into states that belong to the
same or different modes. From the user’s viewpoint, this
3.1 New interface model for the ACC system
To solve the problems with the conventional interface
model described in the previous section, a new interface
model and its user interface had to be designed. In this study,
we assumed that the machine states could not be changed.
Therefore, the standby and active modes were partitioned for
each state to remove the incompatible mode transitions. As a
result, as shown in Fig. 3, the standby mode was separated
into the canceled and armed modes, and the active mode was
separated into the override and active modes. The state and
mode transition table of the new user interface, as shown in
Table 3, indicates that the incompatible mode transitions were
clearly eliminated.
Table 2. Transition among states and modes caused by user-triggered operations in the conventional user interface.
2454
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
Table 3. Transitions of states and modes caused by user-triggered events in the final interface model.
Fig. 3. Final mode design of the interface model for the ACC system.
3.2 New user interface for the ACC system
4.2 Procedure
On the basis of the new interface model proposed in the previous section, we designed a new user interface for the ACC
system, as shown in Table 4. The texts and colors of “ACC”
and “ACC standby,” as well as the color of “Set ** km/h” are
used to identify and distinguish the modes and states more
clearly than the previous user interface. This claim was verified in the experiments described in Sec. 4.
Before the experiment, the participants completed a consent
form and were informed about the goal and experimental
method. In particular, the experimenter explained the ACC
system in detail, including its operational modes and states.
Next, each participant obtained hands-on experience of the
ACC system in the driving simulator.
In the experiment, a driver started the vehicle and turned on
the ACC system. According to the design scenario, a specific
event occurred after the host vehicle arrived at a certain location. The typical events were a sudden maneuver by an adjacent vehicle or the appearance of a road construction sign.
During each event, the driver tried to control the vehicle in a
safe manner by applying the brake or turning the steering
wheel in response to the given situation. The experimenter
observed the subject’s actions and the resulting mode and state
changes in the system. After each event was complete, the
experimenter interrupted the driving simulation to ask the
4. Human-in-the-loop experiments
4.1 Participants
Ten participants, i.e., two females and eight males aged between 24 and 28 years (Mean = 25.2 years, SD = 1.23 years),
were selected from Kookmin University. All had a valid
driver’s license with driving experience of one or more years,
but little experience of the ACC system. They had normal or
corrected to normal vision.
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
subject about the mode and state changes, and their reason for
giving the specific answer. The experimenter noted the answers given in the experiment. The participants were expected
to use the ACC system while driving. Therefore, if the active
mode was canceled by applying the brake, the subject had to
resume the system as soon as the emergency situation was
over. The subjects were given no clues about the correct answers. An experimental session lasted 30 minutes from start to
finish.
4.3 Driving simulator
The experiments were conducted in a fixed-base driving
simulator with TNO PreScan software [17]. Although there
Table 4. New user interface based on the final interface model for the
ACC system [13].
2455
was a driving simulator equipped with a motion platform in
our university [18], it was not used in this work because its
development environment did not allow easy implementation
of different graphic user interfaces for instrument clusters. The
simulator had three 42-inch widescreen displays, thereby creating a 130° horizontal and 25° vertical field of view, as well
as a 10-inch display for the gauge cluster. The input device
was a Logitech G27 racing wheel with gas and brake pedals.
The buttons on the wheel were defined to produce various
operations during ACC mode transitions. In the driving simulator, the ACC system was implemented based on PreScan
using MathWorks Matlab and Simulink. PreScan is a physicsbased simulation platform that provides various sensor emulation functions, which facilitate the rapid development of various active safety systems or advanced driver assistance systems.
4.4 Scenario
To perform tests while driving in the simulator, a road with
six lanes (Three lanes in each direction) was modeled using
PreScan. The road was based on the F-1 race circuit of the
Fig. 4. Test road and the locations where the events occurred.
Table 5. Events and their expected state transitions.
2456
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
Table 6. State and mode confusion rates with the conventional ACC driver interface in the experiments [13].
30
30
25
24
20
15
10
8
M ode Con fu s ion R a te
S ta te Con fu s ion R a te
25
20
15
13
10
5
5
0
0
Conventional Interface
Proposed Interface
(a)
1
Conventional Interface
Proposed Interface
(b)
Fig. 5. Box plots for state and mode confusion rates for the conventional and proposed ACC interfaces: (a) A box plot for state confusion rates; (b)
A box plot for mode confusion rates.
Singapore Grand Prix. As shown in Fig. 4, the road had
straight and curved sections, and its total length was about 5
km. The vehicle started from the lower left corner of the road.
Ten events were designed that occurred at the specific regions
numbered 1–10 in Fig. 4. Table 5 shows the traffic situation,
the expected driver operation, and the ACC state change during each event.
5. Results and discussion
The following data were collected for each event in the experiment: the state and mode before the event, the subject’s
operation, the actual state and mode after the operation, the
state and mode that the subject recognized after the operation,
and the reason why the subject thought they were in a specific
state and mode. We examined the state and mode confusion
rates for the conventional and proposed ACC interfaces. Their
medians and means are shown in the box plots in Fig. 5. We
performed the Wilcoxon signed-rank tests to analyze the effect of interface type because the distribution of the rates are
not normal (p = 0.005 (< 0.05) in the Anderson-Darling normality test) and the sample size is only 10. The medians of the
state confusion rates of the conventional and proposed interfaces are 30% and 0%, respectively, and the medians of the
mode confusion rates are 10% and 0%, respectively. The
means of the state confusion rates of the conventional and
proposed interfaces are 24% (SD = 8.43%) and 8% (SD =
11.35%), respectively, and the means of the mode confusion
rates are 13% (SD = 11.60%) and 1% (SD = 3.16%), respectively. The test results show that p = 0.018 (< 0.05) in the difference of the state confusion rates of two different interfaces,
and p = 0.022 (< 0.05) in the difference of the mode confusion
rates. Therefore, it can be concluded that there is a statistically
significant difference in the state and mode confusion rates
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
2457
Table 7. State and mode confusion rates with the new ACC driver interface in the experiments [13].
between the conventional and proposed interfaces.
Tables 6 and 7 show confusion tables for the conventional
and proposed user interfaces. The confusion tables allow us to
investigate how the participants recognized the changes in the
state and mode of the ACC system during the experiment
because they compare the actual states with the states recognized by the subjects. The data in the shaded diagonal cells
indicate correctly recognized states. As shown in Table 6,
73/100 state changes were recognized correctly using the conventional user interface. Among the 27 cases where state confusion occurred, 16 cases were related to mode confusion. The
highest mode confusion rate occurred in the standby mode.
The standby mode was confused most often with the active
mode. The armed state in the standby mode was confused
with the gap-control or speed-control states in the active mode
at a rate of approximately 40%. These cases of confusion occurred in event 8 when the ACC system was turned off and
then on. In the current ACC system, after the ACC button was
pushed, the state changed into the armed state in the standby
mode. When the set button was pushed again, the state
changed into speed-control in the active mode. However,
some of the participants were confused and they believed that
the state changed immediately into speed-control or gapcontrol in the active mode after the ACC button was pushed.
This type of mode confusion was experienced by most of the
participants who were not familiar with the ACC system.
They had no experience of a standby mode in most electronic
appliances, which usually execute immediately after they are
switched on. Thus, an appropriate ACC system should have
an improved operational flow where the set speed is set and
the speed-control state activates immediately after the ACC
system is turned on by pushing the ACC button.
Table 6 shows that another high rate of mode confusion occurred in the canceled state in the standby mode, which was
confused with the off mode at a rate of 23% and with the gapcontrol state in the active mode at a rate of 32%. These cases
occurred in events 3, 4, and 5 when the subjects pushed the
brake pedal because an adjacent vehicle cut in suddenly, or
when a vehicle in front stopped suddenly. In the cases where
the canceled state was confused with the off state, the subjects
thought that the state became the off state in the off mode
because the system was not working anymore after they
pushed the brake pedal. However, the actual state was the
canceled state in the standby mode. In the cases where the
canceled state was confused with the gap-control state in the
active mode, the subjects thought that even if they pushed a
brake, the state would return to the gap-control state in the
active mode after they released the brake so the ACC system
would continue to work. This mode confusion was also related
to the graphic user interface in the ACC system. When the
state changed from active to canceled, the ACC indicator light
was still turned on and it was the same color. Although the
“Set *** km/h” display showing the set speed disappeared in
this case, the subjects perceived that the speed information
was not related to the operational mode, and thus they believed that the ACC function was still working. Therefore, in
the graphic user interface of the ACC system, a change in
modes needs to be displayed clearly in the screen region related to the mode. Another cause related to this confusion was
that the subjects expected the brake pedal to work in a similar
manner to the gas pedal, but it actually worked differently.
When the gas pedal was pushed in the active mode, the state
changed temporarily to the override state but returned back to
the previous active mode as soon as the gas pedal was released.
However, if the brake pedal was pushed and released in the
active mode, the state did not return to the active mode and it
remained in the canceled state in the standby mode. Therefore,
in an ACC system, it is necessary to reduce the mode confu-
2458
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
sion among users by ensuring that the mode transition works
in the same manner in both cases for the gas pedal and brake
pedal applications.
Table 7 shows a confusion table for the proposed user interface in the ACC system. The table shows that 92/100 state
changes were recognized correctly whereas eight changes
were recognized incorrectly. One of these errors involved
mode confusion where the canceled mode was recognized
incorrectly as the armed mode. A relatively high confusion
rate occurred between speed-control and gap-control states in
the same active mode. The cases where speed-control was
misrecognized as gap-control occurred mainly in event 2,
where the host vehicle was driving in the gap-control state
and the two vehicles in front were in the same lane. Next, a
target vehicle in front left the lane and thus the ACC state
was changed automatically to speed-control by the system.
However, 3/10 subjects recognized the state incorrectly as
gap-control because another vehicle was present in front in
the same lane. By contrast, in event 6, only one driver recognized the state incorrectly because there was no vehicle in
front when the target vehicle left the lane. In all cases, although the speed-control and gap-control states were confused, the driver was not surprised because the two states
belonged to the same mode and any operation caused the
same mode transition.
These low mode confusion rates demonstrate that the proposed driver interface reduces mode confusion because the
same user input operation always leads to the same mode transition and the mode information is displayed clearly in the
mode indication area on the gauge cluster screen.
6. Conclusions
In this study, we analyzed and verified mode confusion using the conventional and our proposed interface model and
user interface in the ACC system based on a formal method,
followed by human-in-the-loop experiments in driving simulators. The results and contributions of this study can be summarized as follows.
The conventional driver interface for an ACC system was
tested using a formal analysis method to determine whether its
cognitive model of users might cause mode confusion. We
also analyzed whether the interface model was consistent with
the standardized operational modes.
Based on the analysis results of the conventional interface,
we developed a new driver interface as an alternative that is
more robust to mode confusion, where we followed the rules
for clear user interface design.
The experimental results showed that the proposed interface
model, which was verified a formal method, was very effective in reducing mode confusion compared with the conventional interface model. The total rate of mode confusion with
the conventional user interface was 16%, whereas that with
the proposed use interface was only 1%. The results also
showed that the formal method is an effective and useful tool
for designing a human–automation interface.
To reduce mode confusion, it is necessary to provide the
driver with a transparent display of the automation state, as
well as correct and succinct information via the user interface.
To prevent mode confusion from occurring, it is necessary to
apply a simple and accurate interface model, as well as providing a clear interface display and internal information in a
transparent manner.
In future research, the following issues need to be addressed.
The ACC user interface proposed in this study prevents
state and mode confusion very effectively, but it has many
modes and they are excessively detailed, which makes it difficult for users to learn and to gain familiarity. This is because
we accepted the conventional machine state model and then
designed the interface model without modifying the machine
model. This strategy restricted the freedom of the interface
model design and prevented us from developing a useroriented system design. Therefore, to obtain a more userfriendly product design, it will be necessary to design the interface model first, before the ACC states are designed and
implemented. The use of fewer modes and consistent mode
transitions would increase the usability of the system.
There is an urgent need to develop optimal user interfaces
for automated systems because of the recent advent of
autonomous vehicles such as Google cars. The National
Highway Traffic Safety Administration (NHTSA) has defined
five levels of vehicle automation. As the level of automation
increases, the role of the driver shifts from primary control to
supervisory control. At level 1 of automation, one or more
specific control functions operate independently of each other.
At level 2 of automation, at least two primary control functions work in unison to relieve the driver of the control of
these functions. At level 3 of automation, the driver can give
up full control of all safety-critical functions in certain traffic
or environmental conditions. Future vehicles will set the level
automatically for a driver or a system, depending on the state
of the driver, vehicle, and environment. Thus, drivers should
not be surprised if they are confused by the levels or modes.
Based on the current NHTSA criteria, the present study only
addressed level 1 of automation. Therefore, it is necessary to
extend this research to level 2 and 3 automated vehicles [19,
20], which include automatic lane centering and lane changing
systems, as well as ACC, and to develop mental models for
the users and driver interfaces that minimize mode confusion.
Recently multi-modal interaction technology has been introduced in driver-vehicle interface [21-25]. In order to reduce
not only driver distraction but also mode confusion and automation surprise, this new technology needs to be investigated
for design of driver-vehicle interface.
Acknowledgment
This research was supported by National Research Foundation of Korea Grant funded by Korea Ministry of Science, ICT
& Future Planning (MSIP) (Grant NRF-2013R1A2A2A0
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
1068766). It was also supported by Research Program 2012 of
Kookmin University in Korea. The authors express their gratitude to Prof. Ji Hyun Yang for providing precious advices and
to Hwisoo Eom for drawing figures and formatting the manuscript.
References
[1] ISO 15622:2010, Intelligent transport systems-Adaptive
cruise control systems - Performance requirements and test
procedures (2010).
[2] D. A. Norman, The problem with automation: Inappropriate
feedback and interaction, not over-automation, Philosophical transactions of the Royal Society of London. Series B,
Biological sciences, 327 (1241) (1990) 585-593.
[3] G. A. Jamieson and K. J. Vicente, Designing effective human-automation-plant interfaces: A control-theoretic perspective, Human Factors: The Journal of the Human Factors and Ergonomics Society, 47 (1) (2005) 12-34.
[4] N. B. Sarter and D. D. Woods, How in the world did we ever
get into that mode mode error and awareness in supervisory
control, Human Factors: The Journal of the Human Factors
and Ergonomics Society, 37 (1) (1995) 5-19.
[5] N. B. Sarter, D. D. Woods and C. E. Billings, Automation
surprise, Handbook of Human Factors and Ergonomics,
Second Ed., Wiley Publishing Company, New York, USA
(1997) 1926-1943.
[6] M. L. Bolton, F. J. Bass and R. I. Siminiceanu, Using formal
verification to evaluate human-automation interaction: A review, IEEE Transactions on Systems, Man and Cybernetics:
Systems, 43 (3) (2013) 488-503.
[7] R. W. Butler, S. P. Miller, J. N. Potts and V. A. Carreno, A
formal methods approach to the analysis of mode confusion,
Proc. of AIAA/IEEE 17th Digital Avionics Systems Conference, Bellevue, WA, USA (1998) C41-1~C41-8.
[8] Y. Horiguchi, R. Fukuju and T. Sawaragi, An estimation
method of possible mode confusion in human work with
automated control systems, Proc. of SICE-ICASE International Joint Conference, Busan, South Korea (2006) 943948.
[9] Y. Horiguchi, R. Fukuju and T. Sawaragi, Differentiation of
input-output relations to facilitate user to input-E international joint conference, with automated control, Proc. of
2007 IEEE International Conference on Systems, Man and
Cybernetics, Montreal, Quebec, Canada (2007) 2570-2575.
[10] H. Furukawa, T. Inagaki, Y. Shiraishi and T. Watanabe, Mode
awareness of a dual-mode adaptive cruise control system, Proc.
of 2003 IEEE International Conference on Systems, Man and
Cybernetics, Washington, DC, USA (2003) 832-837.
[11] M. Heymann and A. Degani, Automated driving aids:
Modeling, analysis, and interface design considerations,
Proc. of the 5th International Conference on Automotive
User Interfaces and Interactive Vehicular Application,
Eindhoven, Netherlands (2013) 142-149.
[12] D. R. Ahn, J. H. Yang and S. H. Lee, A study on mode
2459
confusions in adaptive cruise control systems, Trans. Korean
Soc. Mech. Eng. A. (2015) (Submitted).
[13] S. H. Lee, D. R. Ahn and J. H. Yang, Mode confusion in
driver interfaces for adaptive cruise control systems, Proc.
of the 2014 IEEE International Conference on Systems,
Man, and Cybernetics, San Diego, CA, USA (2014) 41314132.
[14] A. Degani and M. Heymann, Formal verification of human-automation interaction, Human Factors: The Journal
of the Human Factors and Ergonomics Society, 44 (1)
(2002) 28-43.
[15] M. Heymann and A. Degani, Formal analysis and automatic
generation of user interfaces: Approach, methodology, and an
algorithm, Human Factors: The Journal of the Human Factors and Ergonomics Society, 49 (2) (2007) 311-330.
[16] Hyundai Motor Co., http://www.hyundai.com/kr/blu/select
DlExpdList.do#none.
[17] TASS International, PreScan, http://www.tassinternational.
com/prescan.
[18] J.-Y. Lee and W.-S. Lee, A symbolic computation method
for automatic generation of a full vehicle model simulation
code for a driving simulator, Journal of Mechanical Science
and Technology, 19 (1) (2005) 395-402.
[19] T. E. Trimble, R. Bishop, J. F. Morgan and M. Blanco,
Human factors evaluation of level 2 and level 3 automated
driving concepts: Past research, state of automation technology, and emerging system concepts, Report No. DOT HS
812 043, Washington, DC, National Highway Traffic Safety
Administration (2014).
[20] A. Marinik, R. Bishop, V. Fitchett, J. F. Morgan, T. E.
Trimble and M., Blanco, Human factors evaluation of level 2
and level 3 automated driving concepts: Concepts of operation, Report No. DOT HS 812 044, Washington, DC, National Highway Traffic Safety Administration (2014).
[21] C. Pickering, K. Burnham and M. Richardson, A review of
automotive human machine interface technologies and techniques to reduce driver distraction, Proc. of the 2nd IET International Conference on System Safety, London, UK
(2007) 223-228.
[22] B. Pfleging, S. Schneegass and A. Schmidt, Multimodal
interaction in the car: Combining speech and gestures on the
steering wheel, Proc. of the 4th International Conference on
Automotive User Interfaces and Interactive Vehicular Applications, Portsmouth, NH (2012) 155-162.
[23] H. Park, H.-K. Jung and S.-J. Park, Tangible AR interaction
based on fingertip touch using small-sized nonsquare markers, Journal of Computational Design and Engineering, 1 (4)
(2014) 289-297.
[24] Y. Endo, M. Tada and M. Mochimaru, Reconstructing
individual hand models from motion capture data, Journal of
Computational Design and Engineering, 1 (1) (2014) 1-12.
[25] H. Ha and K. Ko, A method for image-based shadow interaction with virtual objects, Journal of Computational Design
and Engineering, 2 (1) (2015) http://dx.doi.org/10.1016/
j.jcde.2014.11.003 (in press).
2460
S. H. Lee and D. R. Ahn / Journal of Mechanical Science and Technology 29 (6) (2015) 2451~2460
Sang Hun Lee is a professor in Automotive Engineering at Kookmin University in Seoul, Korea. He received his
BE, ME, and Ph.D. degrees in Mechanical Design and Production Engineering
from Seoul National University in 1986,
1988, and 1993, respectively. His research interests include computer-aided
design, human-machine interaction, ergonomics, and artificial
intelligence for automotive industry.
Dae Ryong Ahn is a researcher at Korea Automotive Technology Institute in
Cheonan, Korea. He received his MS
degree in Automotive Engineering from
Kookmin University in 2014. His research interests include advanced driver
assistance system, intelligent control
system and human-machine interaction
for automotive industry.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising