- Smart City/Mobility Business HOME
- Search by use case
-
Search by product/service
- event·
seminar -
Case studies and columns
- Inquiry
Revolutionizing autonomous driving development
Digital twin world reconstruction technology
Digital twin technology and simulation technology have become indispensable elements in autonomous driving development.
Traditional development methods require vast amounts of driving data to build highly accurate autonomous driving systems, and when developing new vehicles, dozens of vehicles were actually driven to collect data.
However, simply increasing the amount of data is insufficient to improve the accuracy of autonomous driving. What is needed is an improvement in data quality and the construction of datasets that cover a wide range of scenarios.
On the other hand, in emergency response scenarios such as heavy snowfall or traffic accidents, vehicle verification is difficult and poses safety risks. Digital twin world reconstruction technology and simulation technology are the solutions to this problem.
Digital twin construction service to improve the quality and development speed of autonomous driving
AI development cycle for autonomous driving
There are four processes involved in AI development for autonomous driving.
1. Data Collection
2. Annotation
3. AI Model Learning
4. AI Model Evaluation
This approach involves repeating these four cycles to continuously improve and enhance quality.
Data collection requires the efficient acquisition of high-quality data across a wide range of scenarios. Furthermore, simulations are increasingly being used to supplement missing data during model evaluation.
aiMotive's World Extractor
aiMotive offers World Extractor, which reconstructs digital data from real-world data. Hybrid rendering, combining World Extractor with the simulator "aiSim," enables the reproduction of dynamic objects and the generation of large numbers of scenarios, which are difficult to achieve with 3D reconstruction alone.
Data collection services at Macnica
At Macnica, we can collect data for World Extractor using our vehicle, "macniCAR".
"macniCAR" is equipped with aiMotive's World Extractor and a data collection environment for automatic annotation, enabling us to provide driving data at locations requested by the customer, as well as 3D environment reconstruction data and annotation results.
The value and effectiveness of using digital twins in autonomous driving development.
Features of Hybrid Rendering
Macnica 's original data collection service
FAQ
A. This depends on the sensor setup, calibration, and time synchronization accuracy of the vehicle in question. Additionally, you will need to provide specifications for the data format for data conversion.
A. By providing a license, you will be able to generate a large number of 3D maps. However, you will need to set up the necessary infrastructure.
お問い合わせ
開発支援サービスに関するご質問やご不明点などございましたら、こちらよりお気軽にお問い合わせください。