All News | Boards | Chips | Devices | Software | Archive | About | Contact | Subscribe
Follow LinuxGizmos:
Twitter Facebook Pinterest RSS feed
*   get email updates   *

Mentor tips Azure IoT support and Linux-driven self-driving tech

Apr 24, 2017 — by Eric Brown 675 views

Mentor announced Azure Certified for IoT compliance for Mentor Embedded Linux, and unveiled a Linux-based “DRS360” self-driving car platform.

Mentor, which has long offered its Yocto Project based Mentor Embedded Linux (MEL) commercial Linux distribution, announced new Microsoft Azure cloud support for MEL, as well as its Nucleus RTOS. Earlier this month, its Mentor Automotive unit unveiled a SAE Level 5 compliant autonomous driving platform called DRS360 that integrates multiple sensor inputs in real time. The platform runs Linux on on a Zynq UltraScale+ MPSoC (see farther below).

The DRS360 announcement came only a few days after Siemens announced the completion of its $4.5 billion acquisition of Wilsonville, Oregon based Mentor Graphics. Mentor, which is primarily known for its EDA design tools, is now a business unit called Mentor that is part of Siemens’s product lifecycle management (PLM) software business.


MEL gains Azure certification

The Mentor Embedded Linux development platform, which is based on a Yocto Project distribution, now offers improved IoT connectivity to the Microsoft Azure cloud platform. By joining the Microsoft Azure Certified for Internet of Things program, Mentor enables MEL and Nucleus customers to quickly launch IoT solutions based on pre-certified hardware and software.

MEL architecture
(click image to enlarge)

MEL and Nucleus now ship with an Azure SDK that implements Azure cloud functionality for compliant edge devices or gateways. The SDK lets developers access deeply embedded information like CPU usage, memory availability, power management status, and debugging information, thereby enabling advanced analytics, says Mentor.

The pre-certified hardware profiles “can save time and effort on project specs and RFP processes,” says the new Siemens unit. The extensive list of Azure IoT pre-certified hardware ranges from commercial and community-backed SBCs, including the Intel Edison and Raspberry Pi 3, as well as a wide range of IoT gateways, industrial computers, and signage systems. Microsoft also offers a variety of Azure IoT Starter Kits from Adaruit, Seeed, and Sparkfun based on the Raspberry Pi 3, Edison, Arduino, ESP8266, and Intel NUC.

DRS360 autonomous driving platform

Mentor’s new DRS360 Autonomous Driving platform captures, fuses, and utilizes raw data in real time from a wide range of sensing devices, including radar, LIDAR, and vision systems. The Linux-driven DRS360 platform provides autonomous driving functionality such as sensor fusion and event detection, semantic perception of objects, actuator control, and applications such as situational awareness and path planning. DRS360 “delivers dramatic improvements in latency reduction, sensing accuracy, and overall system efficiency required for SAE Level 5 autonomous vehicles,” says Mentor.

DRS360 fusion architecture
(click image to enlarge)

The DRS360 technology was developed by the Mentor Automotive unit, which claims to be the “number one supplier of automotive Linux.” In 2015, Mentor Automotive delivered a GENIVI-compliant Connected OS platform for advanced driver assistance systems (ADAS) that replaced its earlier Mentor Embedded Automotive Technology Platform (ATP) for in-vehicle infotainment (IVI). The ATP platform in turn was based in part on MEL and on an IVI stack it acquired from MontaVista.

Now, Mentor Automotive has moved to a fully autonomous driving platform that utilizes a “raw data” sensor fusion approach. The technology “eliminates processing at sensor nodes from all sensor nodes” in favor of a central processor that integrates raw sensor data in real time. By eliminating pre-processing microcontrollers from all system sensor nodes, the system offers advantages such as real-time performance, significant reductions in system cost and complexity, and new “access to all captured sensor data for the highest resolution model of the vehicle’s environment and driving conditions,” says Mentor.

DRS360 sensor fusion (left) vs. typical ADAS-oriented solution
(click images to enlarge)

The DRS360 technology provides a streamlined data transport architecture that lowers latency by minimizing physical bus structures, hardware interfaces, and complex, time-triggered Ethernet backbones, says Mentor. The technology is further equipped with optimized signal processing software, advanced algorithms, and compute-optimized neural networks for machine learning.

Although the DRS360 system is said to support a variety of x86 and ARM-based SoCs, as well as a wide array of sensors, it’s initially available on an optimized hardware reference platform that builds on Xilinx’s Zynq UltraScale+ MPSoC. The 16nm FinFET fabricated Zynq UltraScale+, which has appeared in a variety of Linux-driven COMs, is available in quad- and dual-core Cortex-A53 core models clocked to 1.5GHz. The SoCs also offer a Mali-400 MP2 GPU and a Xilinx FPGA that is much more powerful than that found on the similarly hybrid Zynq-7000 SoCs.

The DRS360 hardware platform’s use of neural networking is said to enable fully automated driving within a 100 Watt power envelope. The system is compliant with ISO 26262 ASIL D standards, as well as SAE Level 5 autonomous vehicle specs.

Further information

The Microsoft Azure Certified for Internet of Things support for Mentor Embedded Linux appears to be available now. No availability information was provided for Mentor’s DRS360 self-driving car technology. More information may be found at the Siemens/Mentor Azure IoT announcement and DRS360 product page.

(advertise here)

Print Friendly, PDF & Email

Please comment here...