1. “ Mutual Assistance System
for Automobile Safety ”
Seminar by
Armstrong Ruban.B
2. INTRODUCTION
• Advanced driver assistance systems(ADAS), such as
adaptive cruise control(ACC) systems, are receiving
increased attention throughout the world as a way to avoid
traffic accidents.
• This paper provides evaluation results for ADAS using the
mutual assistance paradigm,in which drivers mutually assist
each other for automobile safety.
• The semi-automatic system was more effective than the fully
automatic system for making positive attitude changes. In
addition, the semi-automatic system was the most effective for
decreasing the driver collision rate.
3. MUTUAL ASSISTANCE
The driver assistance system can be subdivided into the following
three types, adopted from Japanese disaster management:
individual assistance, mutual assistance, and public assistance.
Fig. 1 shows the relation among these three assistance types.
Notification
of braking to
following car
Fig. 1. Diagram of the three types of assistance provided by
driver safety systems. [Adapted from (Kurihashi et al.
4. SYSTEM AND SCENARIOS
• The car driver (recipient) turning right cannot see the oncoming
motorcycle(potential victim), because the view of the oncoming vehicle is
blocked by the truck (assister).
• To prevent such potential accidents, we proposed a mutual assistance
system that would allow the assister (truck) to provide a warning alert to
the recipient regarding the existence of a vehicle (motorcycle) in
his blind spot.
• This system presupposes that charge-coupled device (CCD) cameras (or
millimetre-wave radar) have been installed on the side-view mirrors of
vehicles in order to register the presence of vehicles in the vehicle’s blind
spot.
• Additionally, we assume that the warning is provided by a change in the
assister’s vehicle colour from its actual colour to an unambiguous colour in
the CCD image. Automobile body colour changes can be achieved by
technologies such as Fun-vii provided by TOYOTA.
6. The envisioned merits of this system are as follows:
• The driver as an assister can participate in a safety related
activity by issuing a warning.
• The system can provide mutual assistance
independently of road-to-vehicle or vehicle-to-vehicle
communication
• The system strongly promotes the view that we can help each
other through visual information.
• The system can note the direction of an impending risk
by using real-world information.
7. Scenarios for Evaluation
In this paper, the proposed system is evaluated under four different scenarios.These
scenarios are as follows:
(1). "Collision of a vehicle making a right turn and an oncoming vehicle" (Fig. 3).
The participant plays the recipient or the assister role in the experiment.
Fig. 3. "Collision of a vehicle making a right-turn and an oncoming
vehicle" scenario
8. (2). "Overtaking a bus" (Fig. 4). The participant plays the recipient role in the experiment.
9. (3). "Suddenly rushing out in front of a car" (Fig. 5). The participant plays the assister
role in the experiment.
10. (4). "Avoiding an obstacle" (Fig. 6). The participant plays the recipient and assister roles
simultaneously in the experiment.
11. EXPERIMENTAL ENVIRONMENT
Setup of the driving simulator
• The experiment was conducted in a stationary driving
simulator that uses complex computer graphics to provide a
highly realistic driving environment. The simulated driving
environment is projected on a 120-inch screen in front of the
driver’s seat, and two 100-inch screens project the right-hand
and left-hand side views.
12. Automation level
• Each scenario took place in an urban street route, approximately20-25 min in
duration, and included the following 2 to 4 automation levels.
• Level 1 is a manual warning: the driver can voluntarily provide a warning to a
recipient and its timing is decided manually.
• Level 2 is a semi-automatic warning (small help): the driver can voluntarily provide
a warning to other drivers and its timing is decided semi-automatically.
• Level 3 is a semi-automatic warning (big help): the driver can voluntarily provide a
warning to other drivers and its timing is decided semi-automatically. The difference
between levels 2 and 3 is that level 3 also has a guide feature that advises the driver to
provide a warning. If the driver moves the turn-signal lever at 2.0s or longer before the
collision, the warning is provided at exactly 2.0 s before.
• Level 4 is an automatic warning (big help): The system provides a warning
automatically for other drivers and its timing is decided automatically.
13. Participants
• A total of 15 participants (6 women and 9 men)
participated in this experiment.
• Their driving experience ranged from 1 and 23
years.
• None of the participants were students.
• Informed consent was received from all test
participants.
• Each participant was paid one book coupon worth
1,000 Japanese yen per hour.
14. Hypotheses
The working hypotheses in this experiment are as follows:
• Semi-automatic system (Levels 2 and 3) gives a positive
attitude change for the driver.
• Semi-automatic system (Levels 2 and 3) can restrain the target
level of risk.
15. RESULTS AND DISCUSSIONS
• The inappropriate acceleration ratio for each
automation level. Among the three levels, the inappropriate
acceleration ratios in levels 2 and 4 are higher than that of
level 3.
• Results indicate that the participants did not exhibit risk-taking
behaviour at an automation level of 3.
Therefore, level 3 has a slightly lower target level of risk than
the other levels.
16. • The results also show that collision
ratio is slightly lower in level 3 situation.
17. CONCLUSIONS
Therefore, a semi-automatic system with the
guided feature to advice drivers to provide a
warning is the most favourable automatic level
for the mutual assistance system and has the
potential to restrain risk compensation
behaviour(collision). ADAS are thought
effective measure for reduction of traffic
accident
18. REFERENCES
• Byung, S.Y., Corey W., Greg W., Lokesh H., Sheng-Te T.,Xiaowen W., and Tugrul
U.D. (2014), Planning and Roadmapping Technological Innovations. Chapter7,
Springer International Publishing, Switzerland.
• Trimpop, R.M.(1994), The Psychology of Risk Taking Behaviour.
• Elsevier, North-Holland, Amsterdam.Wilde, G.J.S. (2001), Target Risk 2 -A New
Psychology of Safety and Health-. Chapter 3, Pde Pubns, Toronto.
• Wilde, G.J.S.(1982). The theory of risk homeostasis- Implications for safety and
health. Risk Analysis, 2(4),
209-225.