1. Why is the upgrade of NCAP tests forcing the upgrade of the ADAS verification system?
The evaluation of active safety systems by C-NCAP can effectively reduce the occurrence of traffic accidents and lower the number of casualties in such incidents. The strict requirements of C-NCAP have driven the development of active safety technologies and promoted technological progress in the automotive industry. The evaluation results of C-NCAP can help consumers understand the active safety performance of vehicles and enhance their safety awareness.
On February 12, 2026, China Automotive Technology and Research Center (CATARC) dropped a bombshell – the draft of the “China New Car Assessment Programme (C-NCAP) 2027 Edition” was officially released.
At first glance, this appears to be just a routine upgrade of the regulations. After all, since its first release in 2006, C-NCAP has undergone six revisions. But upon careful reading of this lengthy management rulebook (which is over several pages), a clear signal is being conveyed to the entire industry. The China New Car Assessment Programme is making a third leap from passive safety and active safety to cooperative safety.
The new version clearly stipulates that the final star rating must not only meet the overall score rate requirement, but also must simultaneously meet the minimum score rates set for the three sections. In this new version, the minimum score rate requirement for the active safety section has undergone a qualitative improvement compared to the previous version – the active safety score rate for a five-star rating has increased from 70% to 80%, and for a five-star plus rating, it has increased from 85% to 90%.
| Star Rating | Overall Score Rate | Minimum Score Rate for Occupant Protection | Minimum Score Rate for VRU Protection | Minimum Score Rate for VRU Protection (Pedestrian Protection) | Minimum Score Rate for Active Safety |
|---|---|---|---|---|---|
| 5+ (★★★★★☆) | ≥92% | ≥95% | ≥75% | ≥65% | ≥90% |
| 5 (★★★★★) | ≥83% and <92% | ≥85% | ≥70% | ≥62% | ≥80% |
| 4 (★★★★) | ≥74% and <83% | ≥75% | ≥65% | / | ≥70% |
| 3 (★★★) | ≥65% and <74% | ≥65% | / | / | / |
| 2 (★★) | ≥45% and <65% | ≥60% | / | / | / |
| 1 (★) | <45% | / | / | / | / |
For automotive manufacturers (OEM), Tier 1 suppliers, and third-party testing institutions, the current pressing issue is no longer merely whether the vehicle is equipped with this safety feature, but rather whether this feature possesses high repeatability, high quantifiability, and precise verifiability in extreme standard scenarios. At this point, DAISCH’s ground truth system IFS 3000 is the best solution for verifying reliability.
In recent years, with the continuous updates and iterations of autonomous driving functions, there has been an increasing amount of data from autonomous driving sensors and controllers. Data collection requires the use of high-precision ground truth system to collect data related to autonomous driving. These data can be used to precisely simulate certain traffic conditions in the laboratory, test ADAS perception algorithms and control logic, thereby saving a significant amount of testing and verification time. Without data validation, the algorithms may encounter situations that cannot be judged in actual autonomous driving scenarios, leading to errors. Therefore, reliable and high-quality data collection plays an especially important role in the development process of autonomous driving.
In the automotive industry, especially in ADAS and autonomous driving tests, ground truth system validation has become a relatively mature and widely accepted engineering concept. Whether it is positioning error, heading accuracy, or verifying the performance of perception and decision-making systems, the fundamental premise is to rely on a reference system with sufficient accuracy to objectively quantify the performance of the tested system.
The significance of introducing the Ground Truth System is not to conduct another re-measurement using another set of sensors, but to establish a ground truth with absolute spatial and temporal coordinate clues in the complex physical world. It provides the ultimate basis for resolving all the intricate algorithmic disputes.
Previously, the common ground truth system in the market was the RT3000-GPS/INS inertial navigation combined system from OXTS in the UK. Despite its high cost, due to its excellent performance, there are few alternative products in the domestic market. It was not until 2021 that DAISCH launched the first-generation IFS3000 ground truth system, becoming the only ground truth system device in China with precision comparable to that of Europe and America.
The IFS3000 integrates functions such as multi-constellation GNSS, dual-frequency RTK, high-precision IMU, real-time wireless data communication, networking testing, real-time data processing, and CAN communication interface. It can truthfully record the attitude and kinematic information of the carrier (with centimeter-level accuracy), has the ability to work continuously and normally in harsh environments, and is waterproof. The IFS3000 does not rely on other devices and has excellent operability.
In 2023, the DAISCH IFS3000 DS Insight has undergone further functional upgrades to meet the new challenges of more advanced autonomous driving tests.
3.1 More convenient lane line drawing and tracking
Support user-defined definition of the following four types of V2X measurement items,
- Main vehicle profile to secondary vehicle profile,
- Main vehicle profile to geogeometry,
- Main vehicle sensor to secondary vehicle profile,
- Main vehicle sensor to geogeometry.
By importing the vehicle top view and defining the vehicle dimensions, users can quickly obtain the vehicle outline and edit it by dragging the control points.
The DS Insight software supports importing geographic geometric features such as lane lines and parking Spaces through text or Geojson files
3.2 More intuitive: 4 cameras are displayed simultaneously
The DS Insight software supports users to connect up to four supporting webcams. Users can choose the number of camera window images to be displayed and the corresponding camera for each window image based on their actual application scenarios.
When collecting data, it is necessary to first select the list of logged-in devices and configure the corresponding device for each screen. When CAN data collection begins, real-time data collection from the camera will start simultaneously. The data collected by the camera is synchronously saved with the CAN data, and the playback of the camera data is supported.
3.3 More accurate real-time display of V2V point-to-point distance
Through DLAN communication, the slave vehicle can send information such as location and speed to the master vehicle. The master vehicle can obtain data from the slave vehicle and can calculate the relative data between the two vehicles. Currently, this function can support up to six vehicles in a network. The vehicle profile, positioning output points and other parameters are configured through DS insight to realize the real-time data and image display of the master and slave vehicles and the relative information of the two vehicles.
The IFS 3000 + DS Insight desktop software can help users quickly achieve the tracking and processing of a large amount of dynamic data, especially the acquisition of real-time location, relative distance, positioning status and other data of dual/multiple vehicles, and conveniently monitor and analyze the functions and performance of ADAS or AD systems.
3.4 The functions of DS Insight include but are not limited to:
- AEB testing function
- LKA testing function
- LDW testing function
- Real-time display of point-to-point distances for V2V and V2L
- The lane line display function customizes measurement targets and contours
- Real-time display of collision time
- Dynamic curve display and recording
- Support historical data playback
- Test logs are stored in real time
- Quick setup and loading of custom consoles
- Camera and map connection
The DAISCH IFS3000 ground truth system suite can support customers in fulfilling various scenarios such as product development, verification, and testing.
According to C-NCAP, the application scenarios of IFS3000 include but are not limited to,
- Vehicle Automatic Emergency Braking System (AEB)
- ACC Adaptive cruise control
- TJA traffic congestion assistance
- RAEB rear automatic emergency braking
- HWP highway cruise
- APA Automatic parking
- AVP automatic Valet parking
- Performance verification and development of SLAM technology
- Lane Keeping Assist System (LKA)
All the above scenarios can be realized through the high-precision positioning, attitude measurement and multi-vehicle networking functions of IFS3000. Combined with DS Insight software for data analysis and result verification, the accuracy and scenario requirements of NCAP active safety tests can be met.
5. From the perspective of test engineering, where does the value of IFS3000 lie compared to ordinary positioning and speed measurement devices?
In the development process of high-level intelligent driving and dynamic testing from L2+ to L4, testing equipment is not only a recorder of data, but also an “absolute yardstick” for verifying the boundaries of autonomous driving algorithms. Compared with the traditional positioning and speed measurement devices, there are five dimensions of core engineering value in the DAISCH IFS3000 ground truth system.
- Ultimate precision and high robustness: Construction of “absolute truth” of system trust
- Engineering pain point: In complex operational design domain areas like urban canyons, tree-lined avenues or tunnels where GNSS signal is blocked or even rejected, ordinary equipment is very easy to drift trajectory, resulting in invalidation of test data.
- DAISCH Solution: The IFS3000 adopts the full temperature range and full measurement range calibration process fully self-developed by DAISCH. It is equipped with high precision automotive grade IMU array with zero bias instability as low as 0.7°/h and is deeply and tightly coupled with RTK technology. The system has a horizontal positioning error as low as 0.6cm and a speed measurement accuracy of 0.03m/s. Even in extremely harsh signal environments, it is still able to maintain a highly confident three-dimensional motion trajectory based on a powerful inertial navigation algorithm and thus provides an absolutely safe reference standard for perception and control algorithms.
- 250Hz High-Frequency dynamic response: Eliminating the “time black hole” under high-speed operating conditions.
- Engineering pain points: The testing of extreme conditions such as AEB (Automatic Emergency Braking) and emergency avoidance, there are extremely strict requirements for time delay. The phase delay of low frequency devices may cause distorted evaluation results.
- DAISCH Solution: The IFS3000 enables ultra-high frequency real-time data output up to 250Hz. It can accurately measure the attitude (pitch, roll, yaw) and sudden changes in acceleration of the vehicle at the millisecond level with very low latency. This allows engineers to accurately examine the transient response capabilities of the chassis and control system, and negate any misjudgments that may be made because of a data loss or delay.
- High-dimensional Relative Motion Perception: Empowering Multi-Vehicle Collaboration and Complex Interaction Testing
- Engineering challenges: At present, the intelligent driving tests have developed from single-vehicle intelligence to multi-vehicle competition (such as ACC, FCW, BSD, lane change and overtaking, etc.). Ordinary equipment can not satisfy the high-precision real-time calculation of the relative positions between vehicles in dynamic.
- DAISCH Solution: Overcoming the limits of single point testing, IFS3000 system is equipped with industrial grade high power wireless communication modules, it is able to achieve real-time mobile networking on up to 6 vehicles. The radius of effective communication reaches 1km. It can carry out calculation on the millisecond level synchronously and give the relative distance, relative speed and pre-collision (TTC) time between multiple vehicles in high-speed movement. It is an excellent tool for the construction of vehicle-road collaboration and complex interaction scenario validations.
- Integrated soft-hard analysis engine: From “gathering data” to “finding problems”
- Engineering pain point: A lot of data is generated during the testing process, but the data organization, playback, and the tracing of the problem after testing are very time-consuming.
- DAISCH Solution: Equipped with DAISCH’s specially designed DS Insight data analysis software optimized for autonomous driving engineers, DS Insight supports high-precision trajectory drawing at the centimeter level, precise lane line tracking, and dynamic demonstration of the relative position of multiple targets in bird’s eye view. It converts the boring data stream into easy-to-understand visual reviews allowing engineers to find out Corner Cases at the test spot and reduce the Debug cycle drastically for the algorithm iterations.
- Minimalist Deployment and Efficient Flow: Realizing the brand philosophy of Freeing Up Engineers’ Time
- Engineering pain point: Traditional ground truth system is difficult to install, with many wiring harnesses, and a lengthy calibration process, which is an essential waste of precious time for testing.
- DAISCH Solution: In line with high engineering efficiency, the IFS3000 master-slave equipment has a very consistent hardware. Not only that, it also supports quick and stable installation without disassembly (equipped with dedicated suction cup antennas and in-car brackets), but also one-time vehicle installation and rapid calibration. In case like APA (Automatic Parking), it can import parking space data in one minute. This keeps the cost of test preparation to a minimum, and engineers can really focus on the optimization of algorithms.
In short, ordinary positioning devices just “record positions”, while the DAISCH ground truth system not only records, but also supports high dynamics, multiple targets, panoramic replay, deep comparison and in-depth analysis. The FS3000 is not only the set of high-precision hardware equipment, but also high-efficiency verification methodology provided by DAISCH for the mass application of intelligent driving. It fully guarantees the safety verification of autonomous driving because of its automotive level of reliability and strict requirements on precision with simple engineering experience.
6. Conclusion
The NCAP standards are under revision and the ground truth verification capability is becoming an essential part of ADAS development infrastructure.
From the draft of the C-NCAP 2027 version, we can find that the testing standard of active safety is continuously increasing, with a unprecedented speed. The verification abilities for auto safety have completely left behind the functional demonstration earlier era that was characterized by not being precise.
Behind each of those impressive high-star evaluation certificates is a truly decisive factor – a very precise, repeatable and fully quantifiable testing system. The DAISCH IFS3000 is not just a simple testing hardware accessory, it is an indispensable ground truth reference system in the entire active safety verification chain.
For engineering teams that want to increase the efficiency of ADAS development dramatically, improve the depth of the verification process, and make the absolute persuasiveness of test results, selecting a top-notch positioning ground truth system is not only a tactic optimization, but also a strategic necessity. As the NCAP evaluation heads to a deeper, more difficult territory, the ground truth system is no longer a helpful tool on the test field, but rather is irreversibly becoming a standard part in the ADAS development and verification system.