About the Proving Tests

The research categories and implementation items of the Test Evaluation Department for fiscal 2001 are as follows:

Creation of Test Planning Documents

(1) The assignment of roles and items for measurement are to be determined for each of the proving tests (test course, actual roads, driving simulator) from fiscal 2001 to 2002, and test planning documents for this two-year period are to be created. The result will be creation of test planning documents for each field.

(2) The proving test locations for road section of uninterrupted flow field are to be selected by a survey to identify test locations throughout Japan where the testing would be socially significant. The result has been the selection of seven test locations and the creation of impact scenarios, conceptual views of services, and so on.
Test Implementation

(1) Verification of service effectiveness, acceptance, and design values, and advance testing related to negative checks are to be performed for the seven services on the test course of the National Institute for Land and Infrastructure Management (NILIM) of the Ministry of Land, Infrastructure and Transport. Testing has yielded results showing that implementation of services will change driver behavior toward the safety side.

(2) Dangerous conditions are to be replicated on a driving simulator, and driver reactions to mistaken information and missing information are to be verified. Testing has yielded results showing that the negative safety effect to be expected when information does not provide 100% safety can be alleviated by providing information that encourages cautious driving.

The flow of construction of the Advanced Cruise-Assist Highway System (AHS) for practical application first involves the performance of proving tests and feedback of their results. Decisions about deployment toward practical application can then be made, after which work toward actual deployment will begin. The activities in this fiscal year correspond to FOT. (Figure 1)

Figure 1

Prior to proving tests, advance testing was carried out on the driving course and the driving simulator. The main subjects of testing scheduled for this fiscal year are the advanced safety vehicle (ASV) and AHS. Since AHS is a cooperative vehicle-highway system, evaluations will be carried out from the perspective of the ASV vehicle and from the perspective of the AHS infrastructure. Tests of AHS infrastructure-based systems are also being conducted. At those times, tests of management application are also performed with a view to the effective utilization of services for road management.

Planning for the tests is implemented first by the necessary step of determining what the test is intended to clarify. For example, is the service truly effective? Is the system reliable? Is it too expensive? Will maintenance and management be too great a burden? We are incorporating the views we receive from people with experience or academic qualifications in order to clarify such questions.

One of the evaluation items is whether or not it is appropriate to use a test for verification. Separate studies are carried out of a subject's suitability for evaluation by testing, the cost effectiveness, strategy for promotion of widespread use, and examination of legal issues.

Items that are to be verified by testing include service effectiveness, service acceptance, the validity of design values for infrastructure systems, system functions and performance, and the possibility of effective utilization for road management.

The roles for these items are assigned among the three test fields.

The above has described the overall flow of planning.

In order to determine what is to be evaluated, a hypothetical item for verification is established as shown here. (Figure 2)

Figure 2

An overview of the test planning documents will be provided here. The examples will be the support service for detecting forward standing and slow vehicles on a road section of uninterrupted flow field, and the support service for prevention of crossing collisions in intersection systems. (Figure 3 and Figure 4)

Figure 3

Figure 4

Joint proving tests of ASV and AHS are scheduled to be conducted this fiscal year to verify the cooperative road-vehicle safety cruise assistance systems that function by coordination between the automobile and the road. The joint proving tests will cover the content shown here. (Figure 5)

Figure 5

The rationale for selection of test locations for proving tests on actual roads is (1) to choose locations where introduction of services would be socially significant. Specifically, this means locations where large numbers of accidents occur. Then (2) the locations should be such that the purpose of testing is realized. In other words, they should be locations where the services yield their intended impact by addressing the inferred causes of the accidents that occur at those locations.

Seven locations for proving tests on actual roads have been selected on this basis. (Figure 6)

Figure 6

This shows the overall configuration of the test course system. (Figure 7)

Figure 7

(1) The results from tests of the support service for information on stationary and slow-moving vehicles ahead will be discussed here as an example of the effectiveness of a service. The test procedure is to explain the substance of the service and the human-machine interfaces (HMI) to the test subjects in advance, and to have them drive through once without the service, and then to begin the evaluation testing.

The test results are as described here:

Maximum Deceleration
When drivers suddenly discover the obstacle, they rapidly apply the brakes to decelerate. Their thinking is that reducing the maximum speed will probably act indirectly to increase safety. It is clear here that while the average deceleration without services is 5.5 m/s², with services this is reduced to 3.4 m/s². (Figure 8)

Figure 8
Entry Speed at the Location of Visual Perception of the Obstacle
The question is how fast the vehicle was moving into the location at the point where the obstacle is seen. This was broken down by test subject group, but in any case it appears that there is greater safety with services than without. The differences due to age and sex are not very great, and the effect is clearly apparent overall. (Figure 9)

Figure 9
Brake Initiation Location
This is the distance from the obstacle at the location where the brakes were applied. Without services, the brakes are applied at an average of 179 m before the obstacle. It is apparent that with services the brakes are applied farther in advance, so the vehicle stops more safely. This is before a curve, however, so the effect is not entirely due to the services. Drivers apply the brakes to some extent even without services. (Figure 10)

Figure 10
Maximum Deceleration by Vehicle Following Behind
When the AHS vehicle brakes are applied, the brakes are also applied in the vehicle following behind. It is apparent, therefore, that the maximum deceleration by the vehicle following behind is also shifted toward the safety side by the provision of services. (Figure 11)

Figure 11

(2) The content in verification of design values is verification of the appropriateness of spot communication (DSRC) location. Test subjects were given an explanation of the service content in advance, and took one practice run before carrying out their evaluations. The design values were verified by a questionnaire on the timing of services and by measurement of reaction times and deceleration. The question of how far before a section of poor visibility the DSRC should be installed was approached by calculation using deceleration and driver reaction time figures. Consequently, deceleration and driver reaction times were measured.

The results from tests of design values are shown here. (Figure 12, Figure 13, Figure 14)

Figure 12

Figure 13

Figure 14

There were a total of 30 test subjects, and 80% or more considered the timing of services to be appropriate. The design value for reaction time was set at 5 seconds, and almost all the subjects were within this time. The design value for deceleration was 2 m/s2, and the results showed some instances of even gentler deceleration.

(3) Summary of Testing on the Test Course

The testing showed considerable effectiveness on a road section of uninterrupted flow field, and the conclusion was reached that there would be no problem with test implementation on actual roads. Some aspects were effective in intersection systems while other aspects displayed problems, and it has been decided to carry out further verification. The design values were also mainly in line with expectations. Nearly half of the drivers would look to see the situation, however, so there seems to be a possibility that the installation location could be moved in closer.

(4) Verification Using the Driving Simulator

This involved verification of negative effects due to over-confidence in the services, and verification of the negative effects of missing and mistaken information. The screen used in the simulator had a display much like this. (Figure 15)

Figure 15

The test method was to first conduct the test without services. The obstacle was visible from about 100 m in advance, and the approach speed without services was slightly over 100 km/h. When information was provided in the vehicle, the subjects applied brakes in advance and lowered their speed. The tests were conducted displaying just in-vehicle information, just roadside information, and both in-vehicle and roadside information. Display in the vehicle was more effective than the roadside message signs, and it was found that display both by the roadside and the vehicle was even more effective. (Figure 16)

Figure 16

In order to increase the driving load, the subjects were also made to drive while doing mental sums. The advantage of the driving simulator is that it allows replication of dangerous actions. In this way, it was found that display both by the road and by the vehicle had a greater effect than display in the vehicle alone.

Once the subjects had grown generally accustomed to the test, mistaken information was introduced. When provided information that an obstacle was present, they of course applied the brakes even though actually there was no obstacle. Then they were exposed to missing information. This was the opposite mistaken information, so that, even though an obstacle was present, the display did not show the obstacle but instead simply encouraged cautious driving. This seems to have been slightly more effective than when services were not provided and when curve information was displayed.

(5) Summary of Testing by Driving Simulator

It is possible that the negative effect expected to occur because of missing information can be reduced by providing information to encourage careful driving. In the area of safety and reliability design, more effective safety measures are scheduled to be verified from fiscal 2002.