TI broadens offerings for advanced driver assistance systems with new Vision Software Development Kit, Embedded Vision Engine and DSP libraries

DALLAS, TX -- Texas Instruments announced the availability of its Vision Software Development Kit (SDK) offering developers a flexible framework, comprehensive set of hardware device drivers and a useful set of development tools for more efficient implementations of advanced driver assistance systems (ADAS) on TI's heterogonous architecture. The innovative TI Vision SDK, based on TI's popular SYS/BIOS RTOS, enables multiple algorithms to run concurrently on numerous heterogonous cores and eases integration of new functionality into the system. Additionally, the comprehensive debug and instrumentation features allow algorithm developers to benchmark and profile their algorithms in a system context.

TI is also making libraries available for both its Embedded Vision Engine (EVE) and digital signal processor (DSP) on the TDA2x System-on-Chip (SoC). The libraries include more than 200 optimized functions for both EVE and DSP libraries, providing customers and third parties with the building blocks to jump start development and reduce the time to market. The EVE and DSP libraries are available for low-to-mid and high-level vision processing. At CES 2014, TI showcased a surround view demo and the industry's first 1080p60 dense optical flow in a real-time demo built using the EVE and DSP libraries.

TI's TDA2x family

The TDA2x, incorporating an innovative Vision AccelerationPac, combines an optimal mix of high performance, vision analytics, video, graphics and general purpose processing cores in a low-power envelope, enabling a broad range of ADAS applications scaling from entry to high performance. Furthermore, TI's unique Vision AccelerationPac delivers a more than 8xcompute performance at the same power budget for advanced vision analytics in a more cost-effective footprint. Built on TI's extensive knowledge of algorithms and designed to meet the dynamic needs of the ADAS market, the Vision AccelerationPac contains one or more multiple EVEs, to offer a purpose built ADAS accelerator like no other on the market today.

Robust third party ecosystem

TI's expansion of SoCs and software offerings is driving an increasing number of worldwide ADAS third-party suppliers, offering algorithms, services and integration support on TI's broad portfolio, enabling customers to reduce in-house investment, development time and time to market. Many of those ecosystem players leverage TI's ADAS processors and applaud the launch of the latest TDA2x offering:

"The EVE delivers an impressive performance for a variety of ADAS machine-vision algorithms, such as recognition of traffic signs, vehicles or pedestrians, and structure from motion," said Dr. Michael Stoecker, Director R&D from ADASENS Automotive GmbH.

"The TDA2x SOCs in combination with the SDK allow us to develop next generation ADAS systems which keep up with the growing feature and performance demand for camera based vision systems," said Rene Rollig, senior project manager at IAV Automotive Engineering. "The TI ecosystem allows us to implement algorithms and applications efficiently, which were not feasible before. The new EVE is a powerful companion accelerator, which can be operated independently and in parallel to the DSP. It unfolds its full performance in our target applications especially for frame based image processing. The development tools, libraries and documentation are up to date and our engineers like to use these tools because of the excellent quality."

"With deep technological and automotive industry expertise, Tata Consultancy Services (TCS) provides optimization, integration services and algorithms to Global Automotive OEMs and Tier-1 players on TI's ADAS SoCs including the latest TDA2x," said Regu Ayyaswamy, Global Head, Engineering & Industrial Services, TCS. "Texas Instruments is our strategic partner in ADAS solution development and we look forward to continued development of solutions that markets need most."

"I'm excited by the steep performance increase for each generation of TI's ADAS SoCs," said Felix Eberli, department head embedded and automotive, Supercomputing Systems AG. "This has proven beneficial for a wealth of ADAS algorithms we have ported to TI SoCs for different tier ones and OEMs. A key differentiator is the power-saving EVE cores with high throughput and zero-overhead background data transfers."

Dr. Chao-Jung Chen, President of CSSP Inc. said, "TI has been offering best-in-class SoCs for ADAS space, enabling our team to develop production quality ADAS algorithms such as lane departure warning, forward collision warning and blind spot detection on TI's SoC platforms which are available for TI customers. Moving forward, we are committed to offer our current algorithms on TI's next-gen TDA2x platform to realize multi-functional capabilities on a single chip."

"The EVE in TI's TDA2x offers unparalleled compute performance in a low-power footprint for vision and video analytics algorithms in ADAS applications. Integration of the EVE software tools complete with Code Composer Studio IDE streamlined our workflow, and the examples provided helped us to rapidly leverage the power of EVE in our embedded ADAS applications," says Scott Reardon, President and CTO, D3 Engineering.

ISO 26262 Functional Safety Deliverables

The TDA2x is under development respecting the relevant requirements of ISO 26262 functional safety standard. Supporting safety documentation will be available to customers upon device PPAP and release.

Pricing and availability

TI's TDA2x and associated Vision SDK and libraries are is intended for high-volume automotive manufacturers. Please contact your local TI sales representative for more information and visit http://www.ti.com

Suggested Articles

The world’s largest chipmaker saw a 47% decline in data center sales to enterprise and government, even as it forecast a full year 2020 record of $75B

Working with Jacoti of Belgium, Qualcomm wants to make earbuds recognize the hearing anomalies of users.

Tally upgrade from Simbe Robotics uses Nvidia Jetson GPU for edge processing and Intel RealSense LiDAR for higher resolution images