Skip to content

Here Technologies and AWS unveil AI-powered “SceneXtract” to transform real-world automated driving simulations

Seamless integration of Here HD Live Map and AWS services for accurate real-world simulations, reducing time and costs for automotive developers

Here Technologies, the global leader in mapping and location technology, announces the introduction of SceneXtract, a software tool designed to make it easier and faster for automotive developers to recreate real-world environments for ADAS (Advanced Driver Assistance Systems) and Automated Driving (AD) simulation.

Built in collaboration with Amazon Web Services (AWS), SceneXtract offers an innovative solution for one of the most time-consuming tasks in ADAS and AD development: manually searching, locating, and converting real-world environments into simulation-ready scenes.

SceneXtract leverages the Here HD Live Map, a cloud-based service that enables the automotive industry to access highly precise, accurate and continuously updated map data. The Here HD Live Map contains information such as the topology and geometry of roadways globally, including lane-level attributes and vehicle localization features. The HD Live Map is utilized commercially to help power some of the most advanced automated driving systems available on the road, including Mercedes-Benz’ Drive Pilot and BMW’s Personal Pilot.

SceneXtract utilizes AWS natural language processing (NLP) and generative AI services to help developers quickly locate and export Here HD Live Map data into scenes, in OpenDrive format, for testing, reducing the time and effort typically required for simulation preparation.

Key Benefits of SceneXtract:

  • Accelerated Development: SceneXtract leverages NLP to swiftly search and locate specific map tiles, reducing manual effort and increasing efficiency in scene recreation.
  • Seamless Integration: The tool allows developers to generate and export scene-based data into OpenDrive format with ease and precision, significantly reducing development time and costs.
  • Scalability and Accuracy: By integrating Here’s HD Live Map and AWS’s Gen AI vector search, SceneXtract delivers the highest accuracy for ADAS and AD simulation validation, supporting scalability across various testing environments without concern of AI model hallucinations.

“SceneXtract marks a significant milestone in ADAS and AD development, providing developers with a powerful tool to reduce the time, cost and complexity of simulating real-world scenarios,” said Remco Timmer, Vice President of Product Management at Here Technologies. “By utilizing the Here HD Live Map and AWS’s AI-powered search capabilities, SceneXtract enables more efficient and accurate testing, allowing manufacturers to bring safer, more advanced autonomous driving technologies to market faster.”

Collaboration with AWS

SceneXtract harnesses the power of AWS services, such as Amazon Bedrock and OpenSearch, to offer automotive developers the ability to search, retrieve, and export map data faster than ever. The combination of Here’s world-class mapping capabilities with AWS’s AI technology provides automotive OEMs and Tier 1 suppliers with a scalable and highly efficient solution for their ADAS/AD needs.

“Automotive customers have been increasingly relying on virtual environment simulations to validate safety, convenience, and efficiency features of vehicle ADAS and AD applications,” said Karen Langona, Global Director of Partner Sales, Automotive and Manufacturing at AWS. “We are excited to collaborate with Here in building SceneXtract, a ground-breaking new solution that aims to accelerate innovation and help provide a seamless and efficient way for automotive customers to find scenes and recreate real world-like scenarios for autonomous mobility testing.”

In 2025, Here will continue to develop SceneXtract’s capabilities, including enhanced features and pre-packaged scenes for certain use cases, including SAE L2+ automated systems focused on highways in the U.S. and Germany, and edge cases like complex intersections.

SOURCE: Here

Welcome back , to continue browsing the site, please click here