In system evaluations, we considered what evaluations were appropriate for concept research, system design, and element technologies development. In fiscal 1997, our first goal was establishing evaluation systems and evaluation tools in order to verify the appropriate levels at each step in system development.

As a result, we have clarified the evaluation items, evaluation methods, and evaluation tools which will serve as an evaluation system. For evaluation tools, we have used the settings on the simulator specifications and specification settings for the data collection tools used in the simulator. Furthermore, we have set the policies for the actual equipment test facilities to be used in the future verification tests.

As a process of R&D, we see three major phases of setting user services, system design, and trial systems. We have prepared the cross sectional evaluation criteria for the three phases. Next, we have assumed that the important sections of the evaluation criteria, evaluation methods, required data, and the tools used will change by phase and we have constructed them separately as an evaluation system. At the user service setting phase, the important evaluations are cost benefit analysis and social acceptability. As a result of these, evaluation results such as prioritizing development and further definition of services are expected. Also, at the system design phase, it is important to evaluate from the considerations of verifying system functions and reliability. These results will be utilized in selecting architecture or optimizing system design parameters. Finally, at the trial system construction phase, these technologies or effects will be verified under actual environments.

In systematizing evaluation criteria, we have set the axes of viewpoint, individuals involved, and phases and have organized the evaluation items that arise from their coordinates. For example, we consider the AHS user as an individual involved in the service setting phase and we thought the criteria of accident reduction and traffic congestion relief from the viewpoint of expected results. We organized the structure by considering the fact that for the road manager at the stage of system design, reliability and functionality are very important.
From the axis of viewpoint, we have categorized the components according to its relationship to society when the AHS system is really introduced to our society. Its expected effects were organized by such elements as whether it will affect society, what its relationships with societal resources and costs are, and in technological aspects, whether they will be socially acceptable, what its relationships with existing systems are, and the changes of ideas that changes in society will bring.

Next, we have summarized what methods we use to evaluate these criteria. At the user service setting stage, the evaluation of results is very important. We will use on-paper evaluations, surveys, and simulations to conduct evaluations. Furthermore, since social acceptability is important, surveys or hearings will be conducted for evaluation. Technological opinions are important at the system design stage, so on-paper evaluations or simulations will be used for technological criteria such as functionality and reliability. At the trial system construction phase, we will verify this technology with actual equipment.
Calculating the reduction in accident rates is important as the safety effects which will be the evaluation technique of the service setting phase and it will be done in the following manner. First, accident forms, accident causes, and human elements will be extracted and the categories to be evaluated will be narrowed. Next, evaluation scenarios will be set from these categories and the evaluation parameters to be evaluated will be decided. Based on these, evasion failure rates and hazard evasion failure rates will be calculated using either logical calculations or simulations. Then, the accident reduction rates will be calculated comparing status quo ante to the AHS system combining the overall system reliability, especially the error rate.
At the system design phase, we will verify whether the safety effects targeted at the service setting phase are being achieved by the system design. As a technological criteria, functionality, such as whether they truly respond to pre-set scenarios, and evaluations on their ability to bring the major system design parameters to optimal levels will be verified. As evaluation techniques like these, sensor detection ranges and delays, communication delays, warning timing, and control start timing will be evaluated and optimized through on-paper calculations or through simulations. For comfort and user friendliness, we will use the driving simulator and determine the driving stress of the test subject and its effects on user friendliness. Furthermore, we will analyze reliability based on hazard analysis based on various hypothetical scenarios.

For design evaluation tools, we have in mind accident evasion simulation evaluations and have constructed specifications for the AHS Traffic Simulator for this purpose. As a tool to measure data from actual traffic or actual drivers, we have set up specifications for a human factor evaluation device by altering the JARI Driving Simulator. We also plan to obtain data on ordinary driving action by drivers using survey vehicles. Using mobile measuring devices, we plan to measure the traffic environment, traffic volume, speed, and speed distribution for a certain time period.

For the AHS traffic simulator, we have set specifications to achieve the goal of calculations based on the following input and output conditions. The output evaluation index will be the accident rate. The input will be the system conditions, driver conditions, and external conditions. The system conditions can be set as the parameters of the range of system design parameters and the driver conditions, the operational cycles, response time, acceleration and deceleration rate, and responses upon hazard evasion. The external conditions can be set as various traffic conditions, road composition, and road environments. In order to realize this, we have drawn up simulator specifications which contain modules for AHS function simulation, vehicle and driver simulation, and road conditions.

As one of the data gathering devices, we constructed specifications for a human factor evaluation device. This makes it possible to gather data on hazard evasion in driver hazard evasion operations, response times, response locations, and evasion room. Also, it makes possible the measurement of driving stress and mis-operation rate used in evaluating comfort and user friendliness.

For the second data gathering device, we constructed specifications for conducting measurements in the actual traffic based on the surveyed vehicle. We had a specific driver drive a vehicle and we used various sensors, cameras, heartrate sensors for measuring driving behavior, driver conditions, and vehicle behavior according to the surrounding conditions to gather data.

For the third data gathering device, specifications of a mobile measuring device were set. This was installed at a specific place for several weeks, and data was accumulated by synchronizing the actual traffic flow with the surrounding traffic environment.

At the trial system construction phase, we aim to prove the effects and technology under real life conditions. For criteria, we aim to measure safety, comfort, functionality, and user friendliness under real life conditions. Making several test scenarios, we will evaluate by referring to the simulation data compiled previously and by comparing how these services were performed and what the resulting movements of the vehicle and the driver were.
The real life equipment tests will be conducted using the Public Works Research Institute test course. Network sensors and other equipment will be set up around the course and will be hooked up to the service systems under evaluation. From here, the said services will be performed to the subject vehicles and the multiple vehicles with on board AHS devices and measuring instruments and driver movements will be evaluated. The evaluation criteria will be whether the vehicle movements are achieving the desired target figures, the performance of hazard evasion, degree of reduction of driving stress, user friendliness, etc.


In fiscal 1997, the methods of the above tests included a construction of complete evaluation systems. Within the system, we have selected evaluation criteria and drew up evaluation methods appropriate for each phase. We have also compiled specifications and set up the tools necessary for evaluations. In the future, we will improve the details of this evaluation system and at the same time, we will conduct actual evaluations. The evaluations for safety simulations will be done in conjunction with gathering the necessary actual traffic data. We will also start intense research on the services necessary in order to obtain transportation efficiency such as reducing traffic congestion starting from this fiscal year and establishment of these efficiency evaluation traffic simulators. Lastly, we will continue to implement actual equipment testing tools with the verification tests in mind.