Two of the global auto industry’s leading suppliers are combining forces to develop technology for self-driving cars, aiming to provide manufacturers with a ready-made self-driving platform. Delphi and Mobileye have entered a partnership to build a platform for fully autonomous driving that will be showcased in a matter of months and production-ready for their customers by 2019. By working together, representatives of the two companies say they can rapidly build a system that can be integrated into vehicles of any type and used in urban and highway settings.
“Our two companies are entering a partnership that we believe reshapes the automated-driving landscape,” said Delphi president and CEO Kevin Clark.
The companies did not disclose their levels of investment, but Clark said their development efforts would together add up to “hundreds of millions” of dollars. If they succeed, they could essentially level the playing field, enabling automakers who haven’t yet invested in autonomous systems to catch up quickly.
The partnership is the latest in a series of seismic shifts over the past week in the realm of autonomous driving. Volvo and Uber announced they will put fully autonomous vehicles on the road in test scenarios by the end of this year, and Ford outlined plans to mass-produce fully autonomous vehicles by 2021. As a result, the projected timeframe in which self-driving cars might arrive has rapidly condensed.
Each of these projects describes a different path toward autonomy. Mobileye and Delphi favor an approach that relies more on cameras and radar for sensing the road environment and vehicle surroundings and diminishes the role of lidar for sensing. Mobileye chairman and chief technology officer Amnon Shashua said his company’s platform will use solid-state lidar only as a backup to cameras and radar, which he sees as a more cost-effective approach.
The companies have worked together since 2002 on advanced driving and driving-assist features. Executives see the new partnership as a way to merge complementary technologies they’ve developed individually. Delphi will bring path and motion-planning features and a multi-domain controller that brings sensors together, while Mobileye has developed its EyeQ chip technologies that showcase machine-learning, real-time mapping, and vehicle localization solutions.
Work has already begun on the project, and the executives said they intend to demonstrate the platform in urban and highway driving scenarios during the 2017 CES show in Las Vegas in early January. Fleet testing is expected to begin shortly thereafter, with a production-ready system ready in 2019. Depending on how quickly an OEM worked with the suppliers to integrate the technology with their development plans, Clark said it is possible early adopters could put the systems on the road in production vehicles as early as late 2019 or 2020.
The platform could be customized to be integrated into cars that contain steering wheels, brake pedals, and other inputs for drivers to assume control in an emergency, or be used in cars that don’t allow for any human involvement.
The partnership won’t interfere with or supersede other autonomous projects the companies are pursuing independently. Earlier this month, Delphi said it has partnered with the government of Singapore to develop autonomous cars for on-demand use in urban environments, a project that Delphi expects to result in production of customized vehicles by 2022.
Mobileye has been no slouch. Weeks before it announced, in the wake of a fatal crash, that it will stop supplying Tesla with components for its Autopilot driving-assist system—a decision Shashua said the company has no intention of reversing, by the way—the Israeli company entered into a partnership with BMW and Intel to build and commercialize autonomous cars that are slated to go into production by 2021.
With so many partnerships, there’s certainly a chance for technologies to blur across lines. But Shashua dismissed the notion that this presents conflicts.
“A project of this scale requires as much innovation as possible,” he said. “Multiple attempts, for this technology, is a good thing. It reduces risk and enables us to accelerate this type of innovation.”
from Car and Driver Blog http://ift.tt/2bkr7Bu
via IFTTT
0 comments:
Post a Comment