Technical Features
System-level supplier in the field of Advanced Riding Assistance Systems (ARAS)
SCROLL
Technical Features
System-level supplier in the field of Advanced Riding Assistance Systems (ARAS)
SCROLL
Core Algorithm
Anechoic Chamber
Test Automation
Data Acquisition Tool
Automatic Calibration
IMU Algorithm Compensation
Occlusion Detection
Visual Algorithm
Core Algorithm
HaoHang Technology's self-developed millimeter-wave radar core algorithm framework forms a full-link system covering perception, intelligent processing, functional application and closed-loop verification, delivering precise and reliable support for motorcycle ARAS systems.
Perception & Processing：Radar RF front-end captures electromagnetic signals. After initial target screening and feature extraction, self-developed algorithms including IMU dynamic compensation are fused to suppress interference from vehicle tilting and steering, outputting accurate target range, speed and orientation.
Algorithm Simulation & Control：In Matlab/Simulink simulation environment, core algorithms such as adaptive cruise and collision warning are iteratively optimized. They smoothly convert perception data into control commands and are highly adapted to motorcycle dynamic driving characteristics.
Functional Application & Verification：The algorithm enables ARAS functions including BSD, FCW and ACC. Through scenario tests, data analysis and real-vehicle validation, it ensures stable performance in complex riding scenarios.
Anechoic Chamber
HaoHang Technology is equipped with a professional microwave anechoic chamber, creating a pure performance testing environment for millimeter-wave radars.The wave-absorbing materials covering the chamber can effectively eliminate electromagnetic wave reflections and simulate an ideal interference-free space, providing an accurate and reliable benchmark for radar performance testing.
Based on this professional environment, we place the radar on a precision turntable and, combined with a radar target simulator, verify its core indicators including target detection range, accuracy, resolution, and anti-interference capability.
Test Automation
Full-Link Automotive Electronics Development & Testing System: The Quality Cornerstone from Requirements to Mass Production
We have established a full-process automotive electronics development system covering requirement definition, algorithm development, simulation verification, and vehicle testing.Through rigorous processes and a complete toolchain, we ensure the reliability and safety of every product.
Requirement-Driven: Precisely define the development directionTool Empowerment: Build an efficient development closed loopHierarchical Verification: Ensure quality from module to vehicleAlgorithm Development → Simulation Testing → Use Case Coverage
Through a three-level verification system consisting of Module Test, System Test (MIL/HIL Test), and Vehicle Test,we ensure that every module and every system function performs stably and reliably under real road conditions, safeguarding users'travel.
Data Acquisition Tool
1. The data captured during road testing is transmitted to the playback host computer via WIFI or Ethernet port, allowing the viewing of radar alarms and manually marked problem data during the road test.
2. Manually mark alarm data or correct false/missing alarm data to form reference alarm data, which serves as the ground truth for alarm data comparison.
3. Re-inject the point cloud/target data from the road test into the playback host computer to generate corresponding alarm data, compare it with the ground truth data, and calculate the false/missing alarm rate.
4. For the continuous iteration and optimization of radar software, the optimization effect shall be evaluated by conducting host computer Re-inject testing as a priority.
Automatic Calibration
A fully automatic millimeter-wave radar calibration solution designed exclusively for two-wheeler scenarios, compatible with 77GHz automotive radar, to eliminate detection errors caused by mounting offset, driving vibration and Vehicle attitude changes.
It supports automatic calibration and online dynamic calibration, enabling rapid angle correction, target matching and parameter optimization.
With standardized and high-efficiency calibration, it ensures the safety and reliability of intelligent driving for motorcycles, and facilitates rapid mass production and after-sales deployment.
IMU Algorithm Compensation
Haohang Technology's self-developed IMU dynamic compensation algorithm is deeply optimized specifically for the dynamic riding scenarios of two-wheelers:
Relying on a high-precision Inertial Measurement Unit (IMU), it captures key motorcycle attitude data such as the body roll angle and angular velocity in real time when the motorcycle leaning and steering.
It dynamically corrects the deviation of the radar detection field of view caused by motorcycle body leaning, eliminates the inaccuracy of target recognition due to attitude changes, and ensures that the radar can accurately lock onto targets (core parameters including distance, speed, and azimuth) under any riding attitude.
It provides reliable basic attitude information for the core ARAS functions of front and rear radars, which is used to compensate the alarm function algorithms, ensuring accurate and stable function output, and building a full-scenario active safety barrier for riders.
Occlusion Detection
It provides real-time status monitoring for the intelligent safety riding system of motorcycles. Through the dedicated algorithm of the millimeter-wave radar, it can accurately identify obstructions on the radar surface caused by mud, debris, ice, snow, etc., and immediately output a fault warning signal to ensure that the ARAS functions remain in a reliable working state. This improves riding safety and system stability from the source, leaving no blind spots for safety and enabling smooth and confident riding on every journey.
Visual Algorithm
The visual perception algorithm for the front-view camera is specifically designed for two-wheeler scenario. Based on deep learning and real-time image processing technologies, it accurately identifies road environments, vehicles, pedestrians, and dangerous riding scenarios.
Through efficient computing power optimization and adaptation to two-wheeler working conditions, it stably implements core ARAS functions such as lane keeping, forward collision warning, and automatic high/low beam control.
The algorithm delivers strong robustness under complex road conditions and adaptability to harsh environments including low temperatures, strong light, and nighttime conditions, providing reliable visual perception for intelligent motorcycle riding and comprehensively improving active safety when riding.
