Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Evaluating autonomous functions performance through simulation using interoperable sensor, vehicle, and environment models

Abstract : In this paper we present the simulation software tool Pro-SiVIC for verifying and validating of Advanced Driver Assistance Systems (ADAS) and autonomous (self-driving) vehicles. The specificity of Pro-SiVIC lies in the in-depth virtual prototyping of new or improved perception sensors, and in simulating numerous indoor/outdoor scenarios combining vehicles , obstacles, climate conditions, environment and sensors. Such simulations enable virtual verification and validation of perception algorithms and sensor settings. Pro-SiVIC has been designed to help engineers to build realistic real-life 3D scenarios and experience parts of autonomous functions virtually. As a result, the need for physical prototypes is be reduced and in some cases even eliminated at different development stages from early design to final validation. We will show how the software sensor models can on simulate environmental factors that influence sensor performance such as lighting conditions, weather, and characteristics and behavior of other road users. We will also illustrate how the platform can be connected to real-time data management, to the other tools which use different algorithms of perception, post-processing, data-fusion, path planning and control. The process can be applied so users can quickly and precisely study the performance of embedded systems in both typical and critical use cases, to ensure the product will be safe and reliable in operation.
Complete list of metadatas

Cited literature [33 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02903012
Contributor : Mokrane Hadj-Bachir <>
Submitted on : Thursday, July 30, 2020 - 9:31:51 AM
Last modification on : Friday, July 31, 2020 - 3:13:36 AM

File

HadjBachir072020.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02903012, version 1

Collections

Citation

Mokrane Hadj-Bachir, Philippe de Souza, Javed Shaik, Gruyer Dominique. Evaluating autonomous functions performance through simulation using interoperable sensor, vehicle, and environment models. 2020. ⟨hal-02903012⟩

Share

Metrics

Record views

84

Files downloads

131