According to foreign media reports, if we regard connected and self-driving cars as the next killer app, the data platform will provide the necessary support for connected car applications. To this end, the data platform must meet the following conditions:
Self-driving cars will take into account radar (10-100 KB/s), sonar (10-100 KB/s), GPS (50 KB/s), camera (20-40 MB/s) and more A kind of data output, so that a comprehensive understanding of the location of the vehicle and the surrounding environment of the vehicle.
It has been calculated that the amount of on-board data generated and processed per hour by self-driving cars will be as high as 4 TB. To this end, the data platform needs to support real-time data processing and decision making capabilities (for example, when dealing with braking or accelerating operations).
Industry insiders are increasingly emphasizing the use of machine learning and deep learning techniques to make better decisions. This means that emerging software frameworks such as Caffe2 or TensorFlow are needed to provide support, and more new software frameworks will emerge in the future.
To support these applications, the data platform needs to support a variety of processing engines and data types. It also needs to assist in the complex application development process and ensure that software conflicts are as small as possible.
Distributed computing
In many cases, it is important to implement two-way data transfer between the vehicle and the data center/cloud so that machine learning modeling can be re-scored and improved based on the trials (data) stored in the data center.
The data platform will support connected automotive applications that don't focus on infrastructure agnosTIc, but must support continuous, coordinated data flows for seamless data transfer between data centers and/or clouds and vehicles .
According to industry insiders, the main challenge now is to control the data generated by the sensor. Because of the large volume of data and the extremely fast transmission speed, the data platform needs to use the emerging computing framework such as machine learning to provide real-time processing for such data. A seamless connection between connected cars and data centers.
In the future, data management platform providers will be at the heart of the value creation of connected car ecosystems.
Special Cable,Cable Special,Special Electrical Cables,Special Cables Industries
Shenzhen Bendakang Cables Holding Co., Ltd , https://www.bdkcables.com