Magna MAX4 and Innoviz LiDAR for BMW 2021 production
In what is said to be one of the first serial production contracts for solid-state LiDAR, Magna International and its partner, Innoviz Technologies, will supply the BMW Group with the technology for upcoming autonomous vehicle production platforms starting in 2021. The solution supports up to Level 4 and 5 self-driving systems across multiple vehicle platforms using Innoviz LiDAR in a compact size and at a competitive cost.
Magna says the new business underscores its ability to support customers with flexible, scalable automated-driving solutions and transform innovative technologies into automotive-grade products. While Magna has demonstrated via MAX4 the capability to develop a complete autonomous platform, it can also support customers by providing various building blocks of an overall system as in this case with BMW.
“Our strong systems and integration knowledge and complete-vehicle expertise give us the ability to provide scalable solutions for different levels of autonomy.” said Kelei Shen, President of Magna Electronics.
In August 2017, Magna unveiled the MAX4, which can enable up to Level 4 autonomous driving in both urban and highway environments. The platform contains the sensing and computing building blocks, including radar, LiDAR, cameras, and ultrasonic sensors, as well as an ADAS central computing module, which can be supplied in part or as a whole system to meet customer needs and market readiness.
The LiDAR technology generates a 3D point cloud in real time of the vehicle’s surroundings, even in challenging settings such as direct sunlight, varying weather conditions, and multi-LiDAR environments. In addition, the solution provides a complete computer vision software stack and algorithms to turn 3D vision into critical driving insights. The Magna comprehensive sensing solution delivers on critical needs including performance, cost, and compact size.
“The Innoviz and Magna teams collaborated to meet BMW requirements in quality and validation plans, and functional safety,” said Omer Keilaf, Co-founder and CEO of Innoviz. “Automakers have been looking for a LiDAR technology provider to deliver a mass-market solution, and we are proud to see our product coming to series production.”
Innoviz says that BMW selected its InnovizOne for its unique solid-state, MEMS-based design, as well as for the company’s advanced computer vision capabilities for object detection and classification necessary for Levels 3 through 5 autonomous driving. Innoviz believes the selection is one of the strongest indications yet that solid-state LiDAR is the winner over mechanical spinning and other types of LiDAR technology, supporting last week’s ABI Research report stating: “[S]olid state LiDAR solutions from companies such as Innoviz and LeddarTech have emerged as the LiDAR form factor that will not only help enable robust sensing on autonomous vehicles but also more importantly, satisfy stringent pricing requirements set by OEMs.” And Innoviz says that the announcement is further proof that the industry is on track for series production of autonomous vehicles by 2021.
The InnovizOne offers high angular resolution, high frame rate, an extra-long detection range, and a small footprint—60 mm (2.4 in) tall, 120 mm (4.7 in) wide, and 105 mm (4.1 in) deep. The automotive-grade LiDAR for Levels 3 through 5 autonomous driving will cost in the hundreds of dollars range, depending on volume. InnovizPro, the company’s stand-alone LiDAR, for testing and development of autonomous vehicles, costs in the single-digit thousands of dollars. InnovizPro is available now; the automotive-grade InnovizOne will be available in 2019.
“Another important factor in BMW’s decision was our advanced computer vision capabilities, as we provide a complete computer vision software stack that leverages deep learning and a critical algorithmic layer to turn 3D vision into critical driving insights,” said Keilaf.
Innoviz is developing the software layer that uses LiDAR information to provide object detection and classification such as people, cars, trucks, bikes, driveable area, and lane marking. The computer vision is a LiDAR point cloud data-based product that Innoviz developed to provide its customers a better and faster development process of autonomous capabilities.
“We use unique attributes generated by Innoviz’s LiDAR sensor, harnessed by Innoviz Computer Vision division, to deliver scene perception, calibration, mapping and localization tools for our customers,” said Keilaf. “Innoviz Computer Vision is designed tailored to our sensor’s unique capabilities, therefore it provides an advantage over CV that is applied on other LiDARs.” The LiDAR provides different attributes besides the depth information that gives Innoviz’s technology an edge over traditional approaches.
Innoviz will have to ramp up production to satisfy BMW’s needs. Its current production site in Haifa, Israel, has the capability to produce a few thousand devices per month, but is said to be flexible enough to adjust to higher scale fast. The company is also planning to open a second production line in China starting in Q1 of 2019, allowing for even faster scaling.
Founded in early 2016, backed by $82 million in funding, and now with more than 150 employees, Innoviz has quickly become one of the first LiDAR providers to deliver a complete product to market. The company is backed by strategic partners and investors including Aptiv, Magna, Samsung Catalyst, SoftBank Ventures Korea, 360 Capital Partners, Glory Ventures, and Naver. The majority of its team served in the elite technological unit of the Intelligence Corps of the Israel Defense Forces. Though based in Israel, the company is expanding globally, with offices now in Silicon Valley and plans to open additional offices in Germany and Japan in 2018. Its goals and product roadmap are ambitious.
“LiDAR for autonomous vehicles is just the first step,” said Keilaf. “We are also adding a lot of value in areas of computer vision, HD mapping, SLAM, and more; we can deliver the entire autonomous driving stack. Beyond that, our LiDAR technology can also apply to additional market applications in need of advanced sensing technology, such as mapping, fleets & robo-taxis, security and surveillance, industrial automation, logistics, and more. Right now our focus is on the automotive industry, but we haven’t ruled out other areas for future development.”