- Home
- Technical Products
Enterprise Cloud IT Solutions
Test Measurement
Industrial Measurement
- Solutions
Enterprise Cloud IT Solutions
Test Measurement
- Latest Articles
- About Us
EN
Autonomous Driving Hardware-in-the-Loop Testing Solution
Overcoming Delay and Fidelity Bottlenecks
Implementing real-vehicle-level testing in a virtual environment
An end-to-end autonomous driving simulation solution focused on enhancing simulation efficiency, fidelity, and flexibility. By optimizing the simulation engine and data transmission mechanisms, combined with intelligent driving middleware technology, the solution enables efficient, direct transmission of raw sensor data, eliminating cumbersome hardware interfaces and debugging procedures, thereby significantly shortening development cycles and reducing system costs. Core innovations include an integrated simulation backend that consolidates high-precision models—such as scenarios and vehicle dynamics—into a single high-performance computing node, as well as the introduction of a “dual-mode injection path” to enhance data transmission flexibility and simulation fidelity.
This solution supports everything from rapid algorithm optimization to high-fidelity system-level verification, addressing technical bottlenecks in traditional HiL setups—such as complex link configurations, image distortion, insufficient simulation fidelity, and poor system flexibility—and enabling a seamless transition from early-stage algorithm verification to late-stage system verification.
Program Advantage
End-to-End Simulation of Intelligent Driving Scenarios
Simulates the entire autonomous driving process from sensor input to vehicle control, supporting the creation of complex traffic scenarios (including city streets, highways, and rural roads, among others).
By simulating a variety of environmental conditions and traffic scenarios, it helps R&D and testing personnel test and optimize ADAS/AD systems in a virtual environment, ensuring their reliability and safety in real-world conditions.
High-Fidelity Simulation
The simulation scenarios provided by Hongke closely replicate real-world detection performance and deliver reliable simulation results (the recall rate for multi-object detection algorithms in Hongke’s twin simulation scenarios was 64.68, which is close to the detection results in real-world scenarios).
High-fidelity simulation helps R&D personnel accurately evaluate and optimize autonomous driving systems in a virtual environment, thereby reducing the risks and costs associated with real-world testing.
Compatible with High-Fidelity 3DGS Scene Reconstruction
Compatible with NeRF/3DGS static scenes; supports custom dynamic traffic flows (such as vehicles pulling away at intersections and vehicles following one another on narrow roads).
Developed GGSR (General-Purpose Gaussian Scatter Renderer) to optimize ray-traced sensor rendering in 3D scenes; optimized wide-angle lens rendering to enhance consistency; and supported free camera movement to reduce ghosting (artifacts).
Automated Extraction and Reconstruction of Raw Data Scenarios
Supports the ASAM OpenDRIVE and OpenSCENARIO industry standards and complies with assessments for standard scenarios, regulatory scenarios, and functional safety scenarios, such as those defined by Euro NCAP.
LogSim2WorldSim uses cloud-based automation to process and analyze raw data, enabling slicing, pre-testing, and replay functions based on regulatory scenarios such as Euro NCAP.
Highly flexible and stable sensor data output
Integrates a PCIe-based video feed interface card to enable low-latency, highly stable video data feeding. Through DMA (Direct Memory Access), the latency of data transfer from the GPU to the feed interface card is less than 1 ms, ensuring efficient and real-time image transmission.
It provides 4K-level data channel bandwidth to meet the requirements of autonomous driving systems for high-resolution, low-latency, and highly stable image data, thereby enhancing the accuracy and reliability of HiL testing.
Support for multi-GPU distributed simulation
Supports multi-GPU distributed simulation, leveraging the parallel computing power of multiple GPUs to significantly improve simulation efficiency and performance.
Multi-GPU architectures enable the testing of complex driving scenarios and algorithms in less time, and provide robust technical support for the real-time processing of high-fidelity sensor data and synthetic scenario data.
Application Scenarios
Simulated Data Generation
Integrated Testing Platform for Intelligent Driving
Address core technical bottlenecks such as high costs, long cycles, and low coverage in real-world data collection, and accelerate the optimization and validation of perception algorithms.
Provide an advanced experimental environment that aligns with industry standards, supports the implementation of the full technology stack, and enables the production of high-quality research outcomes.
Closed-Loop HiL Test Bench Case Study
Real-time, scenario-based validation of ADAS/AD systems in a HiL environment
Live Demonstration of 12 Cameras + 6 Millimeter-Wave Radars in Action
- Low-Latency Data Injection Based on DMA
- Integration with NI PXI Devices
- Decentralized Operation
Professional technical support from HONGKEI to help you succeed in your project.
Drawing on its many years of technical expertise in automotive electronics, industrial automation, data security, and life sciences, Hongke provides customers with end-to-end solutions ranging from critical components and precision instruments to system-level platforms.
We don’t just provide products; we’re committed to helping our customers achieve long-term success through professional technical consulting and innovative optimization.
