Data logger aims at autonomous vehicle development
Sensors of self-driving cars such as video cameras, lidar and radar generate data volumes of several terabytes per hour and therefore place the highest demands on a data logger in terms of bandwidth, computing power and storage technology.
Typical applications for the Autera AutoBox are recording large amounts of data during test drives, e.g. during development or for homologation, as well as the playback of recorded data in the laboratory or the prototyping of sensor fusion or perception algorithms in the vehicle. For this purpose, the product has its own storage solution, the Autera Data Storage Unit. In combination with a large-volume Autera Solid State Disk, this provides several terabytes of storage space and a memory bandwidth of up to 50 Gbps.
The AutoBox can be equipped with hardware accelerators such as GPUs or FPGAs to perform computationally intensive tasks such as the development, validation or optimization of AI algorithms. These can be used during data logging to perform intelligent data filtering and preprocessing during the test drive. This procedure saves time when evaluating the data and implicitly increases the storage capacity of the data logging system.
The system provides bus and network support based on the latest standards such as Autosar and Fibex. In order to process the data from the various sensors synchronously, the Autera AutoBox takes time stamps and supports various camera interfaces. A dedicated Autera upload station will be available so that the logged data can be uploaded to a server or cloud infrastructure as quickly and easily as possible. This uploads the recorded data to the data center via interfaces with a very high bandwidth, e.g. 100 Gbit/s Ethernet.
The Autera AutoBox comes with a setup based on RTMaps. With the component-based software development and runtime environment, users can acquire data from different sensors and vehicle buses, time stamp it, synchronize it and play it back. Alternatively, other software environments such as Linux-based applications can be used, for which a documented API of all relevant interfaces is available. In the future, a fleet management solution will also be available for managing and monitoring systems in fleet operations.
More information: https://www.dspace.com/en/pub/home.cfm