Test Sites

This Dataset was created by at the Test Sites:

CARISSMA Outdoor - Ingolstadt, Germany

The CARISSMA outdoor test facilities are as the below image. There is an acceleration zone (grey in Figure 1) with 210m and a dynamic area (blue in Figure 1) with 60m x 70m. The maximum speed is of 100km/h. The THI part of dataset in this task generated, was executed in the acceleration zone of the THI CARISSMA outdoor facilities. This facility also provides rain capabilities by the usage of 20 water sprinklers, such as seen in Figure 2. As there are a limited number of this sprinkles, their disposition changes the water amount in the desired area and grounds illustrates the amount of rain that the facility can provide in relation to the distance on the X-axis of the acceleration zone.

Ground Truth Generation

The ground truth generation was made using the RTK system available at the CARISSMA facility, the measurement method was the usage of the SP80 the RTK system provides accuracy of up to 2cm. This equipment was used for generation locational ground truths

Target Location

In the CARISSMA OUTDOOR proving ground measurements were made at 3 different locations, relative to the sensor 27m, 55m, 82m, and 109m. Figure 5 shows each of the 4 positions. Each position during daytime testing had 3 angles, 0 degrees, 45 degrees, and 90 degrees of target orientation. During Nighttime testing it had only one orientation.

CEREMA PAVIN fog and rain - Clermont Ferrand, France

The Cerema PAVIN (Intelligent Vehicle Platform of Auvergne region) Fog and Rain platform is a unique testing facility capable of generating controlled fog and rain at a significant volume (Figure 6) [3] [4]. This platform is operated by the ITS research team of Cerema, a French public agency attached to the Ministry of Ecological Transition, which guarantees its independence, confidentiality, and neutrality.

The platform is 30 m long, 5.5 m wide and 2.20 m high. This volume allows to reproduce realistic road scenarios, including vehicles, pedestrians, and road equipment such as signs, road markings or other urban furniture elements. These scenarios can be performed under night or day conditions, with or without street lighting. It is also possible to set up reference tests with calibrated targets in reflectance and thermal black bodies. Also, the tests performed in the platform can be associated with complementary measurements with scientific equipment provided by Cerema such as cameras (luminance, SWIR, thermal etc.) or a spectroradiometer.

The platform is mainly dedicated to the automotive (ADAS, AVs, lighting) and road (marking, traffic signs, monitoring) domains but also to other advanced domains such as railway, aeronautics, maritime, construction or military.

Within the chamber, fog and rain of varying intensities can be generated upon request. The fog density replicates the range of Meteorological Optical Range (MOR) from 10m to 1000m, with the Droplet Size Distribution (DSD) accurately representing continental or maritime fogs. Rain conditions can be produced with rainfall rates ranging from 10 to 180mm/h.

The MOR [5] refers to the maximum distance (in meters) at which a calibrated object is visibly distinct from its background. The lower the MOR, the denser the fog. Fog with a MOR of 10m is considered extremely dense and is occasionally encountered on roads. Rain intensity [5] is characterized by the rainfall rate in mm/h, corresponding to the height of water in mm that falls on a surface area of 1m² over a 60-minutes period. The greater the rainfall rate, the heavier the rain. A rainfall rate of 120mm/h corresponds to severe thunderstorm peaks in Europe.

Within the platform, the average rainfall intensity is calculated as follows: the flow of water injected into the pipes is measured at the pump outlet by a flowmeter, and the production area is also known. The ratio of the two gives rain intensity in live. The pump is then controlled to maintain the desired flow rate. Calibration tests were carried out during commissioning and are carried out regularly.

CEREMA target Location

For night-time tests 3 distances were measured, and for daytime only one. Figure 7 shows the position for the night-time tests, where d1=10m, d2=19m, and d3=28m. For daytime tests the Sensor location and target location can be seen in Figure 8, which has only one target p1 at 15m distance from the sensor setup. The daytime was chosen to be in another location due to the nature of the CE proving ground, even on a day setup, with the chamber open, the part that is constructed is gradually.