MENU

SDK speeds deep learning development process for automated driving applications

New Products |
By Christoph Hammerschmidt

Renesas Electronics has unveiled a new software development kit (SDK) for its R-Car family of automotive-centric SoCs. With this, the Japanese chipmaker is launching a complete software platform in a single package. This enables faster and easier software development and validation for intelligent camera and automated driving applications in cars and commercial vehicles.

Automobile manufacturers are increasingly turning to Deep Learning in the development of application software for automated driving. This includes intelligent camera applications and ADAS software. However, according to Renesas’ software experts, most Deep Learning solutions available today are based on consumer or server applications – but these do not operate under the same stringent constraints for functional safety, real-time responsiveness and low power consumption that automotive applications require.

The new R-Car SDK is optimised for use with the state-of-the-art R-Car V3H and R-Car V3M system-on-chips (SoCs). Renesas has developed it for the implementation of rule-based computer vision as well as AI functions for automotive applications. The simulation platform provides both AI and conventional hardware accelerators for highly accurate, real-time simulations. Renesas will continue to optimise this virtual platform, said Naoki Yoshida, vice president, Automotive Digital Products Marketing Division at Renesas. A full set of PC-based development tools will be provided for both Windows and Linux. The tools include several libraries, including support for deep learning, computer vision, video codecs and 3D graphics. In addition to Linux, the new SDK also supports several ASIL-D compliant operating systems, such as QNX, eMCOS and Integrity from Green Hills Software.

For the development of powerful real-time computer vision applications for ADAS and automated driving, Renesas also offers a special version of its e² studio for the R-Car V series. The Eclipse-based open source development environment includes extensive debugging capabilities as well as the state-of-the-art e² studio GUI, which allows users to easily customise and integrate third-party tools. Furthermore, bus monitoring and debug functions for image processing and deep learning subsystems are also supported.

The new R-Car SDK also includes a full set of software examples, common CNN networks, a workshop and application examples. This makes it easier for users who are new to developing with the Renesas R-Car platform to use the technology. At the same time, it enables them to get started quickly with development. The R-Car SDK thus supports the benchmarking of Renesas products and helps developers select the appropriate SoC for their target application.

Thanks to the automatic installer, all software libraries as well as the development environment are quickly deployed on a workstation. Automotive applications designed and developed on a workstation can be ported directly to the embedded development hardware. Last but not least, R-Car partners get access to the R-Car SDK from Renesas within the R-Car Consortium for integration into the roadmap of their R-Car solutions.

The R-Car V-Series and SDK can also be combined with Renesas’ high-performance, low-power RH850 microcontrollers and power management products. This can improve hardware design, reduce validation requirements for functional safety and optimise power consumption for smart cameras and other computer vision-based applications.

https://www.renesas.com/products/automotive-products/automotive-system-chips-socs/r-car-sdk-r-car-software-development-kit

Related articles:

Machine learning boost for complex chip and package reliability analysis

Open platform for software-defined car to speed development

What the Dialog takeover means for Renesas’ automotive customers

Open camera platform speeds ADAS development

 


Share:

Linked Articles
eeNews Automotive
10s