Road condition detection sensors were developed with the capabilities required by AHS services to detect position and speed information on individual vehicles, and based on that information to identify the phenomena of standing vehicles, slow vehicles, and congestion. Three types of sensor were developed to detect vehicles: Visible image type sensor, which utilize images from visible television cameras, infrared image type sensor, which utilize images from infrared cameras, and millimeter-wave radar type sensor, which utilize millimeter waves. Laser type sensor was also developed for detection of pedestrians, bicycles, and so on in intersections.

  Tests of the element technologies in the developed sensors were conducted at the Ashigara Service Area on the Tomei Expressway and the Hiji Junction on the Oita Expressway. This testing verified the basic performance of the sensors. The services were also evaluated on actual roads at such locations as the Maitani District of National Highway No.25 in Nara Prefecture and the Kamiyashiro Junction and Nagoya Nishi Junction on the Higashimeihan Expressway. This confirmed that the sensors satisfy service requirements under certain specific conditions.

  The AHS sensors were developed with the following three objectives:

1. To detect position information and speed information on individual vehicles with a high degree of accuracy.
   
2. To use multiple sensors to monitor service areas that cannot be covered by a single sensor, to integrate the detection results from those various sensors, and to identify such phenomena as standing vehicles; and by means of this continuous service-area integrating function to enable monitoring of traffic over a wide area.
   
3. To develop sensors with varying characteristics in order to enable the selection of sensors suited to the environment where utilized (weather, illumination, etc.) and the services to be applied.

  In order to enable selection of sensors suited to the utilization environment and the applied services, three types of sensor were developed and evaluated: Visible image type sensor, infrared image type sensor, and millimeter-wave radar type sensor.

  The visible image type sensor, infrared image type sensor, and millimeter-wave radar type sensor were subjected to testing of the element technologies in the form of performance testing in a heavy traffic environment at the Ashigara Service Area on the Tomei Expressway and performance testing in a poor visibility environment at Hiji Junction on the Oita Expressway.

  The visible image type sensor yielded stable detection accuracy and detection rates at the Ashigara Service Area, confirming the possibility of their use to detect phenomena for AHS services. Testing at Hiji Junction revealed the detection boundary performance of visible image type sensor in a poor visibility environment.

  Visible image type sensor was deployed on National Highway No.25 (Maitani District) and their detection of standing vehicles due to accident was stabilized under extremely demanding road shape conditions (sharp curves) and camera installation conditions resulting in frequent shadowing (blockage).

  Visible image type sensor uses imagery from the visual television cameras that are widely used for road management and so on and employs image processing technology to isolate vehicles and obtain their position and speed information. In this way they detect standing vehicles, congestion, and other such phenomena.

  There are two typical methods of vehicle isolation: The background discrimination method, which isolates vehicles by means of differences in the background image, and the time subtraction method, which isolates vehicles by comparing images captured at different time intervals.

  The sensors achieved high accuracy, with detection rates exceeding 96% in environments such as that of the element tests conducted at Ashigara Service Area, where the road is largely straight and there is illumination at night. In an environment like that at Hiji Junction, however, where visibility is poor due to fog, sensor performance diminishes with deterioration in visibility.

  The evaluation at the Ashigara Service Area showed that detection performance diminished in heavy rain because the vehicle contrast was reduced when the high-temperature muffler area, which is used as a vehicle characteristic by infrared image type sensor, was cooled by the rain. However, performance of these sensors was found to be stable under normal weather conditions. The evaluation at Hiji Junction also showed that infrared image type sensor stands up well to poor visibility conditions caused by fog.

  Infrared image type sensors at Ashigara Service Area and Hiji Junction were verified to have environmental durability and detection accuracy superior to visible image type sensor, even though both sensors have the same visible image systems.

  Testing of services at the Kamiyashiro Junction and Nagoya Nishi Junction verified high levels of safety and stable visibility under conditions of demanding road shapes (sharp curves and varying grades) and the influence of shadowing.

  Infrared image type sensor identifies vehicles by detecting their far-infrared radiation and conducting image processing. The results are used to derive position and speed information, from which standing vehicles, congestion, and so on are detected.

  These sensors generally recognize vehicles by pairing the background (road surface) with vehicle characteristics (high-temperature points such as muffler and radiator, and low-temperature points).

  Performance was determined to be reduced when rain reduced vehicle temperatures. However, these sensors yielded high detection rates even at night without illumination, and they also achieved high event detection performance (100%) in service tests on the Higashimeihan Expressway.

  Millimeter-wave radar type sensor showed high vehicle detection performance and superior environmental durability in element technology testing at Ashigara Service Area and Hiji Junction. As these are active sensors, they can more easily separate overlapping vehicles (shadowing) than image system sensors. They are also more accurate and are able to detect vehicles even in heavy mist and snowstorms.

  These show great promise for uses as AHS sensors in foggy and cold regions where image sensors cannot be expected to perform adequately.

  Millimeter-wave radar type sensor emits 76-77-GHz millimeter-waves onto the road being monitored then measure the target vehicle's speed from the Doppler shift of the reflected waves and the vehicle's position the lag time of the reflected waves.

  These sensors can detect vehicles in multiple lanes by emitting waves while scanning across the width of the target road. These sensors are capable of detecting vehicles under environmental conditions that make detection difficult for image sensors, even including, for example, dense fog and falling snow (snowstorms). Since they are active sensors, they can positively differentiate vehicles under conditions of overlapping in which image sensors are unable to separate them, so long as the sensors can receive waves reflected from the vehicles. However, since they are installed at a low height, they may be completely unable to separate some overlapping vehicles when there are large numbers of vehicles overlapping (shadowing) in heavy traffic environments.