Applications

Applications using the platform

Applications using Pro-SiVIC

Simulation used in the ANR Safeplatoon R&D project

The SafePlatoon project addressed the solution for platooning in the case of autonomous vehicles. The simulation part in the project focused on the cases which would have been dangerous or difficult to reproduce with the real testing.

In this video, we present the capabilities of the simulation with Pro-SiVIC to generate simulated sensor data for various climatic conditions. During the project, Pro-SiVIC has been used to produce simulated camera or lidar data to test algorithms in heavy degraded conditions.

Simulation of the Quasper R&D Bench

One of the key prototypes of the Quasper R&D projects, the hardware benchmark developed at IFSTTAR/LEMCO for automated ADAS testing, was developed much faster thanks to simulation by Pro-SiVIC. In particular, simulation was used to develop the automation of the vehicle's behaviour (control of a driving robot), in order to guarentee the repeatability of the benchmark scenarios.

In simulation, the benchmark and examples of ADAS were implemented, for comparison with real world tests. A simulation model of the vehicle was integrated in Pro-SiVIC, and so were the vehicle sensors. As in the real world prototype, the data fusion was performed under the RTMaps platform.

Pro-SiVIC simulated the environment, the vehicle model, and the sensors to create the very realistic virtual scenarios allowing to test the data fusion of the automated vehicle.

Using the same perception and algorithm, the vehicle was prototype to evolve in a wider area than the pedestrian bench, allowing to implement use-cases involving other vehicles. The following video show a simulation of such a use-case.

 

Design and validation environment for perception based ADAS

Because ADAS testing and validations requires to study complex and critical situations, adverse climatic conditions can be simulated with Pro-SiVIC.

The following short video illustrates a use case of the validation application by rainy conditions :

 

Critical conditions for camera modelling often implicate specific lighting conditions. Pro-SiVIC can take into account physical parameters of the environment and of the vehicle to simulate such edgy conditions in a realistic manner. For example the simulation of a rear camera, that would be intended for motocycle detection. Notice how difficult if is to distinguish the motorcycle with this camera model until very late, even with human eye.

Several approaches can be adopted to chose the simulation scenarios adapted for the testing. At some point, it can be useful to implement standardized benchmark scenarios for performance assessment. This is an example of implementation of a NHTSA scenario.