检测到并清除了危险字符!
Home> News> Global First: Streamax OEM AEBS Secures UK TfL Certification, Advancing Intelligent L2 Fleet Safety

Global First: Streamax OEM AEBS Secures UK TfL Certification, Advancing Intelligent L2 Fleet Safety

2026 02-03

Following the landmark UK TfL AEB certification—through which Streamax became the first supplier globally to meet the TfL 2.5 safety requirements for public bus applications—this article provides a technical deep dive into the development of the OEM L2 capability framework. Beyond the certification milestone itself, it examines how the system achieves long-cycle operational stability under stringent zero–false-trigger requirements, and how the coordinated integration of AEBS and ACC forms a stable and scalable foundation for Streamax OEM deployment in real-world commercial vehicle environments.

Streamax's world-leading AEBS technology has enabled BYD's double-decker pure electric buses to successfully enter service in London, UK.

OEM AEBS: Key Progress Under the UK TFL Framework

A High-Standard Validation Framework for Bus Operations

TFL2.5_AEB is a mandatory safety technology requirement issued by Transport for London (TfL) for newly introduced public buses. It mandates that all new buses be equipped with an Automatic Emergency Braking (AEB) system designed to continuously monitor the traffic environment ahead and automatically apply braking when a potential collision risk is detected, thereby reducing accident occurrence.

Unlike traditional European regulatory standards, the UK TfL framework is purpose-built for real-world bus operating scenarios and introduces a stringent requirement of zero false triggers over 300,000 operational cycles. This places higher demands on system stability, consistency, and decision-making reliability. Under this validation framework, Streamax's OEM AEBS successfully passed UK TfL certification, completing a comprehensive system-level verification against high-standard application requirements.

Streamax engineering team on site in the UK during AEBS certification testing under the TfL regulatory framework.

System and Engineering Enhancements Supporting High-Standard Validation

In response to the stringent requirements of the UK TfL framework, Streamax's OEM AEBS underwent targeted system-level enhancements. Building on an industry-leading velocity estimation approach, advanced visual multi-cue ranging was introduced, improving Ranging Accuracy and Velocity Fusion Consistency, and enabling more precise control of longitudinal and lateral errors.

The engineering framework supporting Streamax OEM AEBS was upgraded in parallel. By establishing a debugging toolchain centered on the XCP protocol and RiL Simulator, system debugging and issue localization became more efficient, providing strong support for high-frequency validation and rapid iteration. In addition, end-of-line calibration processes in the OEM production workflow were optimized, significantly improving delivery efficiency.

At the same time, the decision and control system underwent a Version 2.0 refactoring, with systematic optimizations focused on reducing false-trigger events. These improvements enable OEM AEBS to meet high-standard validation requirements while also supporting more stable delivery and scalable deployment.

Certification Outcomes and Multi-Platform Validation

Results ultimately demonstrated the system's capabilities. Supported by these technical and engineering advancements, Streamax OEM AEBS successfully obtained UK TfL AEB certification, making Streamax the first supplier globally to complete this certification. During validation, the system delivered stable and reliable overall performance, achieving record-high evaluation results within the TfL AEB framework, with one vehicle platform setting a new benchmark score.

In parallel, OEM AEBS has successfully completed certification and acceptance across multiple vehicle programs, with its capabilities validated on different vehicle platforms, establishing a solid foundation for continued rollout and scalable deployment.

ACC: From Functional Development to the Maturation of Key Capabilities

Beyond AEBS, ACC represents another core capability within the OEM L2 system, complementing safety assurance with stable and controllable longitudinal driving performance.

A Core Longitudinal Control Function within OEM L2

ACC (Adaptive Cruise Control) is a core longitudinal control function within the OEM L2 system, responsible for continuously regulating vehicle speed and following behavior under varying road and traffic conditions to ensure stable and controllable driving performance. Within a development cycle of just three months, the Streamax engineering team achieved a major milestone by advancing ACC from initial implementation to acceptance approval.

Under stringent acceptance requirements—including zero tolerance for issues identified over more than 1,000 kilometers of road testing—the development of ACC presented significant challenges. On the one hand, the complete functional framework had to be built from the ground up and validated within a limited timeframe; on the other hand, the system was required to maintain stable and predictable control behavior despite the inherent limitations of commercial vehicle chassis actuation performance.

Integrated Scenario Handling and Control Stability Enhancements

The functional design of ACC is not limited to a single driving scenario. Instead, it covers a wide range of typical conditions, including following control, cruising control, cut-in and cut-out events, curved-road control, and deceleration for stationary targets. These capabilities are progressively coordinated within a unified control framework, allowing the system to maintain stable and consistent behavior throughout continuous driving.

During the advancement of ACC, Streamax introduced multiple technical innovations. The system underwent comprehensive enhancement around multi-cue ranging capabilities, with the introduction of key features such as Joint Optimization of Lane and Obstacles and Hybrid 2D/3D Velocity Estimation, resulting in improved visual perception accuracy.

At the same time, lane perception capabilities were continuously refined. By extending support for scenarios such as S-shaped curves and lane bifurcations, the system maintains reliable lane understanding in complex road segments.

At the control strategy level, ACC incorporates vision-based anticipatory deceleration, enabling smoother speed regulation without reliance on additional sensors. Combined with multi-model target tracking and IMU state fusion, longitudinal control behavior remains stable and predictable across flat roads, slopes, and varying load conditions. Together, these coordinated technical enhancements form a comprehensive ACC capability set designed for real-world driving environments.

From Validated Capabilities to Scalable OEM L2

The continued advancement of OEM AEBS and ACC forms a critical foundation of the OEM L2 capability framework. With OEM AEBS achieving UK TFL certification and ACC making steady progress from functional development to practical validation, these capabilities together provide stable and practical support across key dimensions such as safety assurance and continuous longitudinal control, enabling reliable baseline performance in real-world road environments.

As capability maturity increases and scenario validation continues to accumulate, OEM L2 will further support the exploration and deployment of higher levels of automated driving.

Q1: How does OEM AEBS ensure stability while minimizing false interventions in real-world fleet operations?

A: OEM AEBS is designed around long-cycle operational consistency rather than single-scenario performance. Through system-level optimization across perception, decision, and control, the solution prioritizes stable judgment under continuous driving conditions. This approach allows the system to respond decisively when needed while avoiding unnecessary interventions during normal operation.

Q2: How does ACC maintain smooth longitudinal control across complex traffic scenarios such as cut-ins, curves, and stop-and-go conditions?

A: ACC is developed to operate across a wide range of real-world driving scenarios within a unified control framework. By coordinating perception inputs with refined control strategies, the system maintains coherent speed regulation and following behavior, even in high-frequency scenario transitions, ensuring smooth and predictable longitudinal performance.

Q3: How do AEBS and ACC work together within the Streamax OEM L2 architecture for fleet applications?

A: Within Streamax's OEM L2 architecture, AEBS and ACC operate as complementary core capabilities at different levels. AEBS is designed for high-risk bus operating scenarios and has been validated under high-standard regulatory frameworks such as the UK TfL requirements, providing reliable safety intervention. ACC, by contrast, forms the general core and foundation of Streamax's OEM L2 system, supporting continuous longitudinal control during normal driving. Working together within the same architecture, ACC provides stable, ongoing driving control, while AEBS delivers a rigorously validated safety fallback in critical situations, collectively balancing safety and usability in OEM L2 applications.

Streamax is committed to the responsible and ethical deployment of technology. Our solutions are developed with a privacy-by-design and security-first architecture. All data processing occurs locally on the edge device, ensuring that personally identifiable information, including biometric data, is neither stored nor transmitted to the cloud, thereby adhering to global data sovereignty regulations.

The AI features and performance metrics referenced in our materials are based on data from extensive internal testing and validation under controlled, laboratory-style scenarios. These results are provided to demonstrate our technological capabilities and direction; however, actual performance may vary in real-world operating environments and should be validated by the end-user.

Our AI models are trained on diverse, legally sourced datasets and are designed to function strictly as decision-support tools for human operators, not as autonomous systems. We actively mitigate algorithmic bias and our development process aligns with emerging global standards for AI ethics and functional safety.