Creation of digital models for accelerated and reliable testing of automated systems in adverse weather DOI

Tuomas Herranen,

Erik Henriksson,

Pak Hung Chan

et al.

Published: Nov. 13, 2024

Automated systems are becoming widespread in many fields, e.g. transportation, exploration, defence, rescue, etc. These need to build a comprehensive and robust situational awareness, detailed terms of spatial temporal resolution. This awareness is based on the data provided by suite perception sensors (e.g. camera, LiDAR, RADAR, etc.). Due internal external noise factors, quality sensor can be heavily compromised.
It impossible test under all possible environmental conditions safety critical cases. To tackle testing complexity speed up procedures, digital twins models environments needed enable accelerated thorough virtual and/or mixed wide variety non-ideal conditions. In order use virtual/mixed properly assess system performance its safety, simulation-to-reality gap needs reduced as much possible, using high-fidelity combination with validated reproduce accurately that real would produce. work discusses development validation two one outdoor indoor facilities, offering rain fog emulation site. By usage high-resolution geo-referenced point clouds images combined photogrammetry 3D modelling, semi-automatic reconstruction material creation process presented. The created models, collection these production trustworthy realistic data. turn, this allows numerous tests executed reliably. hereby described have been developed part EU Horizon ROADVIEW project∗ will made openly available.

Language: Английский

Creation of digital models for accelerated and reliable testing of automated systems in adverse weather DOI

Tuomas Herranen,

Erik Henriksson,

Pak Hung Chan

et al.

Published: Nov. 13, 2024

Automated systems are becoming widespread in many fields, e.g. transportation, exploration, defence, rescue, etc. These need to build a comprehensive and robust situational awareness, detailed terms of spatial temporal resolution. This awareness is based on the data provided by suite perception sensors (e.g. camera, LiDAR, RADAR, etc.). Due internal external noise factors, quality sensor can be heavily compromised.
It impossible test under all possible environmental conditions safety critical cases. To tackle testing complexity speed up procedures, digital twins models environments needed enable accelerated thorough virtual and/or mixed wide variety non-ideal conditions. In order use virtual/mixed properly assess system performance its safety, simulation-to-reality gap needs reduced as much possible, using high-fidelity combination with validated reproduce accurately that real would produce. work discusses development validation two one outdoor indoor facilities, offering rain fog emulation site. By usage high-resolution geo-referenced point clouds images combined photogrammetry 3D modelling, semi-automatic reconstruction material creation process presented. The created models, collection these production trustworthy realistic data. turn, this allows numerous tests executed reliably. hereby described have been developed part EU Horizon ROADVIEW project∗ will made openly available.

Language: Английский

Citations

0