The Autonomous Driving and ADAS segments are undergoing a metamorphosis, driving complex requirements for compute and sensing capabilities. In order to stay on the cutting-edge of this competitive landscape, automotive system-design engineers need to engineer the right computing architectures. FPGAs provide a unique advantage over other silicon solutions and are ideal to meet the evolving requirements in the Autonomous Driving industry.
Key Challenges in Enabling Autonomous Vehicles
|Performance and Efficiency||To enable advanced features and increasing processing requirements with low power consumption|
|Real-Time Processing||To enable data processing and decsion making in real time|
|Safety and Security||To protect against cyberattacks and ensure functional safety of the vehicle|
As the level of driving automation evolves from ADAS L1 to Fully Autonomous Driving L5, the number of sensors required and the need to process sensor data will increase exponentially. Proliferation of sensors used to provide a comprehensive 3D perception of surroundings for both safety and convenience, and adoption of image sensors with higher resolution, pixel depth and frame rate would require multiple communication interfaces and high data bandwidth.
Intel® FPGA can provide the ideal solution that meets the flexible IO and high data rate requirements of these systems. FPGAs can aggregate the data from multiple sensors (with different types of interfaces, data rates etc.) and convert them into a unified format (e.g. MIPI CSI-2) for output to the compute element further down the AD System.
LiDAR sensor units are becoming ubiquitous for the AD applications. Different architectures are emerging ranging from basic signal processing in the LiDAR sensor on the edge to more advanced features such as fusion and machine learning being implemented in the LiDAR unit. Intel® FPGA can provide the flexibility and scalability to address the signal processing, data fusion and complex parallel processing tasks this application requires.
FPGAs are ideal for cryptographic services that enable basic functionality, such as authentication as well as more sophisticated system level security implementations. FPGAs can manage the secure data handling policy for an autonomous driving system which can be reconfigured to adapt to changing requirements.
To build your own security functions, take advantage of the state of the art features on the FPGA including secure boot, secure key storage, cryptographic acceleration, secure lock, unique device ID, secure debug, and physical tamper detection and protection.