Visteon introduces DriveCore autonomous driving platform at CES 2018
Visteon Corporation unveiled its DriveCore autonomous driving platform at CES 2018. Designed to accelerate the development and commercialization of autonomous driving technology, DriveCore allows automakers to build autonomous driving solutions quickly and in an open collaboration model.
DriveCore was designed as a complete technology platform, consisting of the hardware, in-vehicle middleware, and PC-based software toolset needed to develop machine learning algorithms for autonomous driving applications of Level 3 and above. DriveCore aims to provide automakers a fail-safe domain controller, with a high degree of computing power scalability, which supports the integration of data from multiple camera, LiDAR, and radar sensors.
"To meet the computing demands of level 3-plus solutions, autonomous systems will need highly scalable levels of processing power and the ability to perform sensor fusion across multiple radar, camera, and LiDAR sensors," explained Sachin Lawande, President and CEO of Visteon. "DriveCore is the first open platform in the industry that offers highly scalable computing power and software to perform late sensor fusion to enable rapid development of these autonomous systems."
DriveCore consists of three primary components:
- Compute is a modular, scalable computing hardware platform that can be adapted to all levels of automated driving. It is designed to deliver from 500 gigaflops to 20 teraflops of processing power (with existing systems on chip) in a scalable manner, independent of the type of central processing unit used. It will support Nvidia, Freescale, Qualcomm, and, later, other processor types.
- Runtime is in-vehicle middleware that provides a secure framework to enable applications and algorithms to communicate in a real-time, high-performance environment. It enables sensor fusion in a sensor-independent manner, so sensors can be upgraded as new capabilities become available, such as radar going from 2D to 3D.
- Studio is a PC-based development environment that enables automakers to create and support an ecosystem of algorithm developers. Studio is intended to allow easy integration of third-party algorithms and access to real-life sensor data—complemented by a simulation, validation, and benchmarking environment for algorithms ranging from object detection to camera-based lane detection.
Visteon also announced four new technology partners as part of the DriveCore introduction:
- DeepScale uses deep learning to create an integrated model of the environment in real time, from any combination of sensors. Then, DeepScale's deep neural networks (DNNs) add the perception capabilities needed for automated driving.
- STEER provides a fully automated parking solution, in which the car drops off passengers at a designated point, drives itself to the nearest automated parking zone, waits for a "summon signal," then drives to the designated pick-up point.
- StradVision deploys machine learning algorithms to build advanced object detection and recognition software.
- Automotive Artificial Intelligence provides a graphical simulation environment, offering intelligent traffic and traffic scenarios; it runs in conjunction with DriveCore Studio.
Visteon is hosting demonstrations of DriveCore at its booth.