(Announcement) Dynamic Map Platform to Host Joint Seminar with NVIDIA and MathWorks – Introducing Data Generation for Physical AI Learning and Validation Using High-Precision 3D Map Data and Simulation

Dynamic Map Platform Co., Ltd. (Headquarters: Shibuya-ku, Tokyo; President and CEO: Shuichi Yoshimura; hereinafter "the Company") will jointly host an online seminar titled "Frontiers of Physical AI Learning and Validation Enabled by HD Maps and Simulation – The Data Generation Ecosystem Envisioned by Dynamic Map Platform × NVIDIA × MathWorks" on Wednesday, July 15, 2026, in collaboration with NVIDIA Corporation (hereinafter "NVIDIA") and MathWorks Japan (hereinafter "MathWorks").

3D model image generated using RoadRunner with high-precision 3D map data

With the advancement of Physical AI and end-to-end (E2E) autonomous driving, the development of autonomous driving and Advanced Driver Assistance Systems (ADAS) has become increasingly sophisticated and complex, elevating the importance of data used for AI training, evaluation, and validation. While data generation using 3D simulation is highly anticipated for its scalability, certain use cases still face challenges due to the discrepancy—known as domain gap—between simulated and real-world imagery.

This seminar will introduce the latest trends and potential applications in data generation through a workflow integrating the Company’s HD Maps (high-precision 3D map data), MathWorks’ RoadRunner, and NVIDIA Cosmos Transfer. Specifically, it will explain the end-to-end process of creating 3D scenes and scenarios using high-precision 3D map data and generating high-quality video data suitable for AI training and evaluation.

From the Company, Mr. Kensuke Mageshita, Head of Software Technology Division, and Mr. Isaya Fukudome, from the Automotive Business Division, will present on leveraging high-precision 3D map data as the foundation of this workflow.

The Company has been collecting and developing high-precision 3D map data for use in autonomous driving and ADAS across 26 countries worldwide, including Japan, North America, Europe, Korea, and the Middle East. This session will demonstrate methods for efficiently importing this data into RoadRunner to generate diverse road network environments, along with its potential applications in simulation and machine learning.

Additionally, the seminar will cover an overview and use cases of the Company’s AI training and evaluation dataset, "Raw Scene Pack," as well as a method for constructing semantically rich digital twin environments suitable for AI learning and evaluation by combining the Raw Scene Pack with 3D Gaussian Splatting technology.

The initiatives introduced in this seminar are positioned as advanced use cases under the Company’s "Data for AI" initiative, which aims to leverage high-precision 3D data for AI training, evaluation, and validation.

Moving forward, the Company will continue to promote the use of high-precision 3D data in AI and simulation domains, contributing to the development and validation of Physical AI, including autonomous driving and ADAS.

<Seminar Overview>

Title

Frontiers of Physical AI Learning and Validation Enabled by HD Maps and Simulation – The Data Generation Ecosystem Envisioned by Dynamic Map Platform × NVIDIA × MathWorks

Date and Time

July 15, 2026 (Wed), 14:00–15:40 (JST)

Format

Online

Participation Fee

Free

Organizers

NVIDIA Corporation, MathWorks Japan, Dynamic Map Platform Co., Ltd.

Target Audience

Engineers involved in autonomous driving algorithm development and validation, ADAS perception algorithm developers, professionals interested in OpenDRIVE/OpenSCENARIO, and those exploring 3D simulation environments

Details and Registration: https://jp.mathworks.com/company/events/webinars/upcoming/physical-ai-training-and-validation-enabled-by-hd-maps-and-simulation-5295750.html?s_eid=PRP_1778485934

<About Dynamic Map Platform Co., Ltd.>

Dynamic Map Platform Co., Ltd. was established with backing from the Japanese government and investment from ten major domestic automakers. Headquartered in Japan, the company operates bases in North America, Europe, the Middle East, and Korea, currently conducting business in 26 countries. It provides high-precision 3D data for a wide range of applications, including autonomous driving and ADAS, simulator environment construction, infrastructure management, and snow removal support.

Guided by the vision of "Modeling the Earth"—digitizing the planet—the company serves as a high-precision 3D data platform, co-creating innovation across various industries.

Established: June 2016

Headquarters: Shibuya-ku, Tokyo

Representative: Shuichi Yoshimura

Business: Provision of high-precision 3D data for autonomous driving, ADAS, and diverse industrial applications

Official Website: https://www.dynamic-maps.co.jp/

Official X: https://x.com/dynamic_maps