The autonomous driving controller is an intelligent computing platform for unmanned systems, specially built for the automatic driving of connected and intelligent vehicles. With its heterogeneous distributed hardware architecture and robust functional performance, the asynchronous driving controller as the principal hardware carrier facilitating vehicles to achieve high-level autonomous driving, is able to provide the key technical supports necessary for the safe and stable operation of vehicles under complex road conditions. Therefore, it has substantial application potentials and excellent industry adaptability in the field of autonomous driving hardware.
The core heterogenous hardware architectural configuration
The architecture is the basis of the platform's capabilities and functional fulfillment, it features exact solutions for power of computation and safety that are required for autonomous driving. It merges 2 NVIDIA Orin computing units and 1 automotive-grade ASIL-D compliant NXP TC397 processor which all three units cooperate in the following way:
NVIDIA Orin Computing Units: They are the main high-performance computing cores that provide strong computing power for large-scale data processing and complex algorithm operation;
NXP TC397 Processor: With the automotive-grade ASIL-D certification (the highest functional safety level for automotive electronics), it is the one that guarantees safe control of the execution of the critical control functions, thus it balances high computing power and high safety.
The platform AI computing power is also as high as 508 Tops, which means that it can very well handle computation-heavy tasks such as on-the-fly multi-sensor data processing and complicated environment perception. Thus it offers a solid computing basis for high-level autonomous driving without the need for the vehicle to worry about decision delays and perception deviations, thereby increasing the vehicle's ability to react to and manage complicated traffic situations.
The platform uses a modular assembly method, thus it is highly flexible and scalable. By combining modules, the platform can be transformed into different hardware platforms with a varied range of functionalities and performance, thus it is no longer limited to fixed functions of traditional hardware platforms.
This product is designed to be able to cover 95% of the scenarios in which an autonomous driving control can be implemented:
Urban road scenarios: Congestion following, intersection passage;
Highway scenarios: Constant-speed cruising, automatic lane changing;
Complex park scenarios: Low-speed autonomous driving tasks.
By using hardware modules to satisfy the differentiated requirements, the company greatly enhances the industry adaptability and application scope of the product while at the same time decreasing the hardware development cost and period.
The platform is endowed with full-process computing capabilities that include multi-sensor fusion perception, path planning, trajectory prediction, and driving decision-making:
1.Multi-Sensor Fusion Perception: It uses data from lidar, cameras, and other devices and removes the blind spots and mistakes through algorithms to achieve the most accurate and complete environmental perception;
2. Path Planning: Determines the safest and most efficient routes based on the data acquired from the environment and the navigation information;
3. Trajectory Prediction: Predicts the future movement trajectories of the traffic participants (for instance, pedestrians, other vehicles) so as to be able to avoid the potential risks;
4. Driving Decision-Making: Returns the most appropriate driving decision by combining the results obtained from perception, planning, and prediction modules.
At the same time, there is a time-sensitive networking (TSN) subsystem built into the controller which can provide time synchronization information to processors and sensor modules on its own. This resolves the issue of time synchronization among multiple devices, guarantees exact time synchronization of perception, computing, and control systems, avoids data misalignment and command delays, and, therefore, leads to the enhancement of system stability and reliability to a higher degree.
The platform is a model of interface compatibility as it supports a variety of interfaces such as 5G, Ethernet, 10-channel CAN, and RS232/485:
5G Interface: Ensures fast and low-latency wireless communication, thus allowing real-time data interchange between vehicle-cloud, vehicle-vehicle, and vehicle-road and consequently facilitating vehicle-road collaboration and cloud-enabled autonomous driving;
Ethernet Interface: Is the solution to high-bandwidth data transmission requirements within the platform, e.g., sensor's data in high-definition can be transmitted simultaneously from multiple sources;
10-Channel CAN Interface: Is capable of establishing a connection to in-vehicle CAN bus devices with the help of which the real-time collection of the vehicle status information and precise issuing of control commands can be achieved;
RS232/485 Interface: Maintains togetherness with the old devices, raising platform harmony and scalability, and adjusting to the different vehicle hardware ecosystems.
FAQ – Frequently Asked Questions
1.When and where is the Expo going to be held?
The Expo is going to be held from May 13 to 15, 2026, at Hall C, Xiamen International Conference and Exhibition Center (XICEC), Xiamen, China.
2. How big is the exhibition?
The event spreads over 40,000 m², featuring 350+ companies, and it is expected that over 30,000 professional visitors will come from all over the world.
3. What are the program activities?
There will be over 80 professional forums and events that will delve into topics such as smart mobility, transportation communication, safety, and sustainable development.
4. How many countries and regions is it related?
Participants from more than 80 countries and regions will attend, thus it is a global meeting for the innovation of intelligent transportation industry. ">
5. Are there possibilities for cooperation?
Yes. The event with over 1,000 global partners provides a wide range of opportunities for business cooperation, technology exchange, and investment.
6. Who can I contact for the details?
Kindly connect with the Organizing Committee via the official website's Contact Us section for additional details.
