Some of the new products on show
Open-architecture platform to assess autonomous driving stacks
Klas
At the inaugural ADAS & Autonomous Vehicle Technology Expo in San Jose, Klas will showcase RAVEN, an open-architecture platform of computing, storage, networking and connectivity for in-vehicle development and assessment of autonomous driving stacks.
The challenge for many AV developers is that legacy logging systems lack the power to log and process the terabytes of raw data that come from integrating GMSL-based cameras and automotive Ethernet-based sensors such as lidar.
To collect and assess vehicle data efficiently and effectively, AV developers need additional computing for visibility, to ensure that the data set collected is fit for purpose; filtering, for pre-processing and tagging of relevant data sets; compression, for ease of offloading terabytes of data in the shortest time; and encryption, for secure storage and transfer of raw data.
RAVEN is a modular system that meets the needs of modern AV development by providing scalable compute platforms to log, process and securely transfer data. Furthermore, it dramatically simplifies access to data for use in hardware-in-the-loop (HIL) or software-in-the-loop (SIL) environments.
The holistic platform lightens the load for AV development by providing all the necessary computing, datalogging, data storage, networking and connectivity in a compact, standardized and ruggedized form factor designed for life on the road.
AV developers can easily integrate compute capabilities into existing test harnesses or replace legacy logging systems by selecting the standalone RAVEN logging and storage module, the TRX D8.
In San Jose, Klas will showcase a range of RAVEN modules that include scalable x86 compute (up to 96 cores) to identify and filter valuable data sets; GPU, where data sets can be assessed, compressed or encrypted with Nvidia Turing; logging via the capture and storage up to 240TB of data in a swappable storage cassette; connection to CAN, Ethernet and automotive Ethernet vehicle networks; connectivity, where data can be securely accessed live over wi-fi, LTE or 5G from remote locations.
In addition to RAVEN, Klas will showcase its Blackrock virtualization, automation and management software suite. Blackrock is designed to simplify laboratory integration of agile software development environments with real-world driving scenarios running on RAVEN.
Booth: 2050