Method and apparatus for data driven simulation for autonomous vehicle training and testing

The method generates simulated sensor data using a network-based simulation platform to address slow development cycles in autonomous vehicle training, enabling efficient and adaptable data generation for diverse scenarios.

WO2026129023A1PCT designated stage Publication Date: 2026-06-25LEDDARTECH HOLDINGS INC
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
LEDDARTECH HOLDINGS INC
Filing Date
2025-10-28
Publication Date
2026-06-25

AI Technical Summary

Technical Problem

Current methods for training autonomous vehicle virtual drivers rely on hand-engineered technology and manual processes, resulting in slow development cycles and inadequate generalization to new geographies and unseen scenarios, necessitating improved data generation methods.

Method used

A computer-implemented method using a network-based simulation platform to generate simulated sensor data based on real-world sensor data, allowing users to select and define scenarios through a user interface, replicating sensor output formats and object information for autonomous vehicles.

Benefits of technology

Facilitates efficient and customizable data generation for training virtual drivers, enhancing the ability to handle diverse scenarios and improve the reliability of autonomous vehicle systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CA2025051422_25062026_PF_FP_ABST
    Figure CA2025051422_25062026_PF_FP_ABST
Patent Text Reader

Abstract

A computer-implemented method, a device, and a computer readable storage medium are disclosed for generating simulated sensor data based on real world sensor data using a network- based simulation platform. The simulated sensor data may be used to train and / or test a virtual driver for an autonomous vehicle. The method is performed at a computing device with access to a user interface. The method includes: exposing, via the user interface, sensor data-selection user interface elements for selecting real world sensor data to provide to a simulation server; exposing, via the user interface, scenario definition user interface elements for defining a simulation scenario that is a variant of a real world scenario in which the real world sensor data was collected; providing information regarding the simulation scenario to the simulation server; and receiving, from the simulation server, simulated sensor data generated by the simulation server according to the simulation scenario.
Need to check novelty before this filing date? Find Prior Art