ACU
Ecotrons Automated-driving Control Unit (ACU) is designed to be the core controller of autonomous vehicles or ADAS vehicles. It is based on NVIDIA Xavier processor, and can process variant vision sensor data and handle the sensor fusion. It is designed as automotive grade embedded controller, compare to those bulky industry computers. Further more, it includes a ASIL-D rated micro-controller, like NXP MPC5744. Combining the super computation power of Xavier and a ASIL-D micro together, Ecotrons ACU offers the production intent autonomous vehicle controllers, which could meet the requirements of both high performance computation and the functional safety, the biggest concern of AVs.
Ecotrons ACU can read data from variant sensors, e.g. multiple cameras, radar, LiDAR and the cloud data; it has enough ports for 6 cameras, Ethernet input from LiDAR, CAN bus, Flexray for radar, and other data steam. It can be integrated with V2X data system, and 5G data network. It can be used for L2 to L4 level autonomous vehicles.
ADAS Controller
Designed for L2 or L2.5 Automated -driving
Ecotrons ADAS controller is based on NXP S32V234 processor. It can also process the vision sensor data, from multiple cameras, LiDAR, Rader, etc. It can be used for L2 or L2.5 level ADAS system. Ecotrons ADAS controller is capable of AEB (automatic emergency brake), ACC (adaptive cruise control), APA (Automatic Parking Assistance), LDW (Lane Departure Warning), etc. Ecotrons ADAS controller can also come with a ASIL-D rated micro on board, at the customer’s option, to solve the functional safety concern.
Various Types of Communication Ports Reserved for Cutting-edge Sensors
Ecotrons ACU provides multiple types of communication ports that meet the requirements for most of the autonomous driving sensors available on market nowadays. LVDS channels of the ACU can be used for vision-based applications. The support of CAN and Ethernet communications provides capabilities for radar, LiDAR and all other CAN or Ethernet based sensors, obtaining a more accurate and reliable driver assistance system.
Powerful Processors
Ecotrons ACU deploys two powerful main processors – NVIDIA Xavier and NXP S32V234. S32V234 MPU is NXP 2nd generation vision processor and a member of 32-bit Arm® Cortex®-A53 S32V processor family. Xavier is NVIDIA’s Tegra series processor designed for mobile devices, which includes Tegra X1, Tegra X2 and Xavier.
| EAXVA01/Xavier | EAS2A01/S32V234 |
|---|---|
 |
 |
Xavier is NVIDIA’s Tegra series processor designed for mobile devices, including Tegra X1, Tegra X2 and Xavier. The system-on-chip includes six processors: ISP (Image Signal Processor), VPU (Video Processing Unit), PVA (Programmable Visual Accelerator), DLA (Deep Learning Accelerator), CUDA GPU, and CPU. It can perform 40 trillion calculations or 30 trillion deep learning per second. The internal architecture is shown in the figure above, which mainly includes the following calculation units:
|
S32V234 MPU contains an Image Signal Processor (ISP), a powerful 3D Graphic Processor Unit (GPU) and dual APEX-2 vision accelerators, which provides automotive grade reliability, functional safety and security capabilities for supporting computation-intensive ADAS, NCAP front camera, object detection and recognition, surround view, automotive and industrial image processing, machine learning and sensor-fusion applications. S32V234 MPU is NXP 2nd generation vision processor and a member of 32-bit Arm® Cortex®-A53 S32V processor family.
|
 |
 |
Port Resource Description
| Interface | EAXVA01 | EAS2A01 | Description |
|---|---|---|---|
| Ethernet | 1 | 1 | LiDAR input |
| CAN | 2 | 2 | Radar/Ultrasonic radar input |
| FAKRA | 6 | 6 | HD camera input |
| RS232 | 2 | 2 | IMU input & Debug interface |
| HDMI | 1 | 1 | Image display |
| USB Type-C | 1 | – | Debug interface |
Applications
