Quick Takeaways
  • Horizon Robotics showcased advanced driver assistance capabilities of the iCAR V27 using its HSD platform and Journey 6P chip.
  • The system operates without HD maps, enabling autonomous roaming mode through end-to-end AI architecture.

The Horizon Robotics HSD solution was showcased during an iCAR technology event where Beijing Horizon Information Technology Co., Ltd. demonstrated the advanced driver assistance capabilities of the new iCAR V27 SUV production model. The vehicle integrates the company’s Horizon SuperDrive platform to deliver enhanced perception, decision-making, and autonomous driving support through a high-performance computing and sensor architecture.

High-Performance Computing Platform

The iCAR V27 SUV is equipped with the Journey 6P automotive chip, delivering computing power of 560 TOPS to support complex autonomous driving workloads. This high-performance processing capability allows the vehicle to analyze large volumes of environmental data in real time while maintaining responsiveness and reliability for advanced driver assistance functions.

The computing platform processes input from an extensive sensor suite to provide comprehensive environmental awareness, enabling the system to understand surrounding traffic conditions, road features, and obstacles.

Multi-Sensor Perception System

The vehicle integrates a total of 27 high-precision sensors designed to provide robust perception coverage across multiple driving scenarios. The sensor configuration includes LiDAR along with other perception technologies that collectively enhance the system’s ability to detect objects, measure distance, and track dynamic road participants.

This multi-sensor approach improves detection accuracy and redundancy, allowing the system to maintain stable operation in complex or unfamiliar driving environments.

One-Stage End-to-End Architecture

The intelligent driving system adopts a one-stage end-to-end architecture that strengthens scenario generalization capabilities. Unlike conventional approaches that require multiple processing stages, the unified architecture allows the system to directly transform sensor inputs into driving decisions.

This design improves learning efficiency and adaptability, enabling the platform to manage diverse real-world situations without relying heavily on predefined rules or detailed environmental mapping.

HD Map-Free Operation

A key capability of the system is its ability to operate without HD maps or city-specific adaptation. Traditional autonomous driving systems often depend on detailed high-definition maps that must be prepared and updated for each location. In contrast, the architecture used in the iCAR V27 enables broader deployment by reducing reliance on map infrastructure.

Adaptive Environmental Understanding

The intelligent driving platform can quickly interpret surroundings even when encountering unfamiliar scenarios. By continuously analyzing sensor inputs and contextual driving information, the system can select appropriate driving strategies in real time.

Autonomous Roaming Mode Capability

The vehicle introduces a roaming mode that allows autonomous route planning without fixed navigation paths. Instead of depending on predefined routes, the system can dynamically plan movements based on the environment and traffic conditions.

This capability expands the operational flexibility of intelligent driving systems and represents an important step toward scalable autonomous mobility solutions in future passenger vehicles.

The demonstration highlighted how advanced computing, sensor fusion, and AI-based decision-making can work together to support next-generation driver assistance technologies in production vehicles.

Company Press Release

Click above to visit the official source.

Share: