English

"New Window" of Car Lidar Track

149
0
See translation
2023-01-09

The cost reduction logic of the vehicle mounted lidar is driven by the core supply chain in addition to the acceleration of the front loading mass production scale. This is also regarded as the core competitiveness of the new cycle of lidar.

 

Lumotive is a start-up company invested by Bill Gates, the founder of Microsoft, which provides laser radar companies with core components based on beam steering technology. It uses liquid crystal super surface (LCM) and silicon manufacturing processes to greatly improve manufacturing efficiency.

 

In the past few years, several lidar companies have turned to innovative solid-state solutions based on MEMS micro mirrors or optical phased arrays. However, due to the small optical aperture of the MEMS micro mirror and the low efficiency of the optical phased array, both of these schemes lack real high performance.

LuMotive's scheme is groundbreaking, based on the unique LCOS customization process, to convert the semiconductor chip into a "dynamic display" and manipulate the laser beam deflection by using the optical bending characteristics of metamaterials. LCM chip does not contain moving parts. It adopts mature semiconductor manufacturing process and LCOS technology, and has the advantages of low cost, high reliability, small size, flexibility and integration (for example, integration with headlights).

 

This week, LuMotive announced that it had obtained a new round of financing led by a venture capital company under the Samsung Group. At the same time, the company confirmed that at present, more than 24 companies have cooperated with it to develop new laser radar solutions using its unique chip design.

 

In the past few years, Lidar has also accelerated the process of chip self research.

Last November, Yijin Technology, a vehicle specification level MEMS lidar solution provider, completed a round C financing of tens of millions of dollars. Accelerating the self research progress of core chips became the company's next new goal, with the ultimate goal of reducing the cost of comprehensive BOM (bill of materials).

 

"Self research of chips is the core competitiveness," said Li Yifan, CEO of Hesai Technology. Just like in the era of fuel cars, well-known car companies will certainly research engines by themselves. The use of self-developed chips can effectively improve product quality, reduce costs, simplify the supply chain, and optimize performance.

 

At present, the semi-solid lidar AT128 with the largest shipment of Hesai is equipped with the second generation chip developed by ourselves. Based on pure solid state electronic scanning technology (E-Scanning), it realizes the array integration of 128 groups of laser receiving channels, greatly improving the product integration.

 

The FT120 blinding laser radar launched last year is equipped with a third-generation chip. Compared with AT128, the single chip of FT120 integrates an array composed of tens of thousands of laser receiving channels, achieving higher point cloud density in a smaller volume.

 

"FMCW technology will benefit from Intel's silicon photonics manufacturing expertise, thus further reducing costs." At the end of 2020, Mobileye announced the 2025 autonomous vehicle sensor system development plan, a self-developed laser radar.

However, the current cost of lidar is not the real starting point of scale. From $1000 to $500, when we can go down to $300 or even less than $100 will be a decisive turning point.

 

The beam steering module of LuMotive, together with SPAD array and low-power VCSEL, can form an efficient and complete combination scheme of laser radar chips.

"In terms of reducing the size and cost of the lidar, the breakthrough of the chip level lidar system will help significantly reduce the cost of the sensor portfolio to achieve fully automatic driving." Amnon Shashua, CEO of Mobileye, predicted that the time point would be around 2025.

 

In the opinion of the insiders, "the core electronic components of the lidar are being integrated into special integrated circuits, which have the advantages of higher density, lower cost and higher reliability. This trend roughly follows the Moore's Law of integrated circuits, which means that it is possible to significantly reduce the volume, weight and cost of the lidar."

 

For example, Mobileye uses Intel's 3D packaging technology to integrate CMOS circuits with silicon photons. "This integration is the key to providing performance and cost optimization of optical transceivers." At the same time, by integrating silicon photonics modules with computing resources, it can break the disadvantage of using more I/O leads, so as to achieve lower power consumption, greater throughput between computing units, and fewer pins.

 

The company previously disclosed that the cost of each lidar SoC is expected to be around hundreds of dollars, an order of magnitude lower than the cost of the current mass production system. What's more, depending on the own wafer factory of Intel, the parent company, you can take the cost initiative in your own hands.

 

This means that the competition in the lidar industry is gradually extending from the early application and market innovation to the real supply chain technology innovation, and the threshold of competition in the next cycle of the industry is looming.

 

According to the monitoring data of Gaogong Intelligent Automobile Research Institute, from January to November 2022, 98400 standard lidars will be equipped on the front of passenger cars in the Chinese market (excluding imports and exports). Since this year, as BYD, Chang'an, FAW Hongqi, SAIC, GAC, Chidu, Lutes and other car enterprises have entered the laser radar boarding cycle, they will continue to drive the market growth.

 

Among them, the models facing the private consumer market will mainly be configured with one forward facing and two blind areas, while the B-end front installation of L3/L4 will mainly be configured with one forward facing and four blind areas. It is expected that the delivery of standard equipped lidars will reach the scale of 2 million lidars per year by 2025.

 

At the same time, the laser radar is also one of the key elements for the technical added value and pricing breakthrough of vehicle models. At the same time, depending on the additional perception capabilities provided by the laser radar, high-level intelligent driving can be shortened, especially the landing time of urban NOA.

 

Source: Gaogong Intelligent Automobile

C
    • No comments
    Collect
    Report