Dynamic Map Platform Co., Ltd. (Head Office: Shibuya-ku, Tokyo; CEO & President: Shuichi Yoshimura; “Dynamic Map Platform”) announces that Ryoto Miyake of the Software Technology Department, Technology Unit, will speak at the online event”ADAS, AV, and AI Meetup”, hosted by Voxel51 on September 17, 2026 (Pacific Time).
Voxel51 is a U.S.-based company that provides FiftyOne, an open-source platform for computer vision. The company hosts community events around the world for engineers and researchers working in AI, machine learning (ML), and computer vision. This event will showcase the latest technology trends and use cases in advanced driver assistance systems (ADAS) and autonomous driving.
Miyake will deliver a presentation titled “From Survey-Grade Maps to Physical AI: Scaling Real-World Data for Training and Simulation”, introducing Dynamic Map Platform’s initiatives to leverage high-precision 3D data originally developed for autonomous driving applications to support Physical AI development.
The presentation will highlight Dynamic Map Platform’s efforts to develop data for AI training, evaluation, and simulation by utilizing image data, road surface models, and 3D Gaussian Splatting (3DGS) data based on the company’s high-precision 3D data assets covering approximately 1.8 million kilometers worldwide. Dynamic Map Platform will also explain methods for structuring and enhancing real-world spatial data for AI development, as well as the value of large-scale datasets that combine geometric accuracy, semantic information, and diverse real-world environments.
In recent years, the importance of real-world data used for training, evaluation, and validation has increased significantly in the Physical AI field, including autonomous driving and robotics. Leveraging the expertise gained through the development and provision of high-precision 3D data for autonomous driving, Dynamic Map Platform is advancing its “Data for AI” initiative.
In addition, Dynamic Map Platform’s AI-native sample dataset, released on Hugging Face in June 2026, has recently become available through the Voxel51 Dataset Zoo, a dataset catalog provided by Voxel51. During the event, Voxel51 is expected to demonstrate this sample dataset, showcasing its usage and potential to the AI researcher and developer community.
[Reference Release]
- September 9, 2026
“Dynamic Map Platform Expands Data Availability for the AI Research Community; AI-Native Dataset Now Available Through the Voxel51 Dataset Zoo”
https://www.dynamic-maps.co.jp/en/news/news-865/
Through its Data for AI initiative, Dynamic Map Platform will continue to expand the expertise gained from developing and providing high-precision 3D data for autonomous driving into the AI domain. By strengthening collaboration with researchers and developers worldwide, the company aims to provide the data foundation that supports the social implementation of Physical AI.
＜Event Overview＞
| Event Name | ADAS, AV, and AI Meetup |
| Date & Time | September 17, 2026, 9:00 AM – 11:00 AM (PDT) September 18, 2026, 2:00 AM – 4:00 AM (JST) |
| Format | Online (Zoom) |
| Participation Fee | Free |
| Presentation Title | From Survey-Grade Maps to Physical AI: Scaling Real-World Data for Training and Simulation |
| Speaker | Ryoto Miyake (Software Technology Department, Technology Unit) |
| Organizer | Voxel51 |
| Event Information & Registration | https://voxel51.com/events/adas-av-and-ai-meetup-september-17-2026 |
