LiDAR to start large-scale mass production — An interview with RoboSense
LiDAR technology is steadily expanding beyond its traditional applications of topography and factory automation into automotive, logistics, and smart infrastructure. According to Yole Développement’s LiDAR for Automotive and Industrial Applications 2021 report, the market for LiDAR in automotive and industrial applications is expected to reach US$5.7 billion in 2026 from US$1.8 billion in 2020, representing a 21% Compound Annual Growth Rate (CAGR).
Founded in 2014, RoboSense is a leading player in this LiDAR revolution with significant activity in new technologies, industrial collaborations, and mass production. Among others, RoboSense has developed a unique scanning technology using a proprietary MEMS LiDAR SoC and is using SiPM for higher sensitivity. RoboSense has also established various collaborations in the automotive industry and for industrial applications. Finally, RoboSense is set to start large-scale mass production of LiDAR, leading to unit cost reduction.
Pierrick Boulay and Alexis Debray from Yole had a chance to discuss with Mark Qiu, co-founder and COO of RoboSense, to learn all the details about the most recent activities of RoboSense.
Pierrick Boulay (PB): Could you please introduce RoboSense?
Mark Qiu (MQ): RoboSense (Suteng Innovation Technology Co., Ltd.) is a world-leading provider of Smart LiDAR Sensor Systems. With a complete portfolio of LiDAR sensors, AI perception, and IC chipsets, RoboSense enhances conventional 3D LiDAR sensors with comprehensive data analysis and interpretation systems. Its mission is to innovate outstanding hardware and AI capabilities to create smart solutions that enable robots, including autonomous vehicles, to have perception capabilities superior to humans.
Founded in 2014, RoboSense is headquartered in Shenzhen, China, and has 5 global branches, including Silicon Valley and the Stuttgart R&D office. As of 2020, it has developed more than 600 patents globally. Its product technology has been widely applied to autonomous and assisted-driving passenger cars, commercial vehicles, automated logistics vehicles, robots, robotaxis, robotrucks, robobusses, new smart transportation infrastructure, and other subdivisions. Among them, the pre-installed fixed-point mass production projects cover supercars, coupes, and SUVs, heavy-duty trucks, and other types of vehicles.
PB: Can you briefly introduce RoboSense’s LiDAR technology?
MQ: RoboSense develops hardware technology and perception software technology, innovatively integrated on a System-on-Chip (SoC) — making its LiDAR system more compact and eliminating mechanical-related failures. This disruptive product subverts the traditional definition of LiDAR as an information collector and transforms it into an intelligent interpreter.
With a market-oriented strategy, RoboSense provides customers with various Smart LiDAR perception system solutions, including MEMS LiDAR, mechanical LiDAR, fusion hardware, and perception software.
Alexis Debray (AD): RoboSense is operating worldwide. We see many projects in many applications related to LiDAR in China. Is there something different about China’s interest in LiDAR compared to other regions?
MQ: The demand for LiDAR in the Chinese market has grown rapidly for applications in automotive, robotic, smart city, and industrial projects. In terms of preference, there are no obvious differences between the Chinese market and other markets in the world. The difference in demand for LiDAR mainly comes from applications in differentiated fields, such as performance and operating conditions in the automotive and industrial fields.
AD: RoboSense is involved in various LiDAR applications, including automotive, smart city, and industrial. What application do you expect to have the most impact on the LiDAR industry in the future?
MQ: As we have seen, each field where LiDAR is applied is a huge market. As various fields gradually begin deploying LiDAR on a large scale, LiDAR products inevitably start appearing in more segments based on the needs of each field.
In the past two years, LiDAR’s mass production has advanced, thanks to the mass production of pre-installed vehicles in the auto industry and the large-scale deployment of autonomous logistics vehicles — two trends that first appeared in the Chinese market.
PB: RoboSense is using and developing different LiDAR technologies (mechanical, MEMS scanner, optical phased array (OPA).
Are these technologies for different applications? What is the roadmap and time-to-market for these technologies?
MQ: RoboSense’s current LiDAR products in the market include a first-generation mechanical LiDAR and second-generation automotive-grade smart solid-state LiDAR (MEMS). These products currently meet the present and near-future market demands.
At present, M1, our second-generation automotive-grade intelligent solid-state RS-LiDAR-M1, has been delivered for mass production projects. We are continuing to upgrade and develop our products based on the MEMS technology platform and plan to introduce two new products with estimated launches in 2022 and 2023.
Looking forward, we are continuing to develop other cutting-edge LiDAR technologies. The roadmap and time-to-market of the next product will be disclosed when technology and market conditions permit.
AD: One key parameter preventing the mass adoption of LiDAR is its cost. What are the main strategies to reduce the cost of LiDAR in the next five years?
MQ: LiDAR applications need to achieve two goals in order to reduce costs: technological breakthroughs, and large-scale mass production.
RoboSense’s second-generation automotive-grade smart solid-state LiDAR M1 is based on two-dimensional MEMS smart chip scanning technology. It is automotive-grade, high-performance, low-cost, and manufacturable — advantages that completely break through the limitations of first-generation mechanical LiDAR as a technological achievement.
In June of this year, M1 was officially mass-produced and delivered to North American luxury new energy car companies, achieving the goal of large-scale mass production.
Therefore, RoboSense has reduced the cost of LiDAR enough to be applied to large-scale projects in automotive and other industries.
PB: How did you choose your MEMS mirror and what were its advantages compared to other devices you tested?
MQ: The MEMS galvanometer used in the RS-LiDAR-M1 is independently developed by RoboSense based on automotive LiDAR requirements, and it fully meets the needs of vehicle-level operations.
The second-generation smart LiDAR architecture based on two-dimensional MEMS galvanometers has clear advantages — a small number of components, a simplified internal structure, and low material and manufacturing costs.
As a MEMS LiDAR model, the RS-LiDAR-M1 has already earned a large number of fixed-point orders from auto companies for its advantages of high performance, low cost, vehicle-class stability, and small size. Also unique to M1 is its “GAZE” function, which dynamically adjusts the LiDAR resolution or refresh rate based on different automated driving scenarios.
AD: You are using SiPMs as receivers in your LiDAR. Could you explain why you have chosen SiPMs and not APDs or SPADs?
MQ: SiPM is a new type of receiver, which improves APD sensitivity by ten to the third power, effectively improving LiDAR’s ranging ability. SiPMs can avoid the problems SPADs incur due to crosstalk: snow-like noise will appear when encountering strong light, and overexposure can cause blindness.
PB: RoboSense recently partnered with Webasto to integrate a LiDAR in a smart roof module (HERE). When would you expect OEMs to implement such types of modules in their car? What use case would be enabled by this LiDAR position?
MQ: The cooperation between RoboSense and Webasto focuses on providing a more compact LiDAR deployment solution for OEMs. In fact, OEMs have already released mass-production solutions for deploying LiDAR on the vehicle’s roof, such as SAIC IM L7 and WM ET7.
Compared to deploying on the air intake grille, deploying LiDAR on the roof means fewer obstructions near the front of the car, a wider view, and avoiding the impact of dirt and sand.
AD: Is there anything else you would like to add?
MQ: The era of mass-produced LiDAR has arrived, marking the tipping point for its large-scale application. RoboSense’s second-generation automotive-grade smart solid-state LiDAR — the RS-LiDAR-M1 based on two-dimensional MEMS smart chip scanning technology — achieved automotive-grade mass-production orders and delivery in June 2021 to North American new energy car companies in large quantities. RoboSense also has two fixed-point project SOPs this year.
Mark Qiu is the co-founder and COO of RoboSense responsible for the company’s strategic management and business operations. He brings extensive experience and insights into sensors, the automotive industry, and autonomous driving to the company.
As a serial entrepreneur, Qiu’s career spans sales, marketing, and operations, covering fields such as consulting management and the Internet. He has been included in the Forbes Asia 2018 30 Under 30 List, Forbes China 2018 30 Under 30 List, and Hurun 30 Under 30 Entrepreneur Leaders in 2019.
Prior to RoboSense, Qiu worked in renowned management consulting companies such as Ogilvy China and Bozhong Enterprise Management Consulting. He has a long history of providing marketing consulting services to Chinese enterprises, leveraging his enriched understanding of business management.
Alexis Debray, Ph.D., is a Senior Analyst at Yole Développement (Yole), dedicated to the production of technology & market reports and custom consulting projects in the fields of Photonics, Sensing, and Semiconductors.
Before joining Yole, Alexis spent 17 years in Japan. He worked for 2 years developing expertise in MEMS technologies and then for 15 years at Canon Inc. as a research engineer, where he contributed to numerous developmental projects focused on MEMS devices, lingual prehension, and terahertz imaging devices.
Alexis is the author of various scientific publications and patents. He graduated from ENSICAEN (France) and was awarded a Ph.D. in applied acoustics.
Pierrick Boulay is a Senior Technology & Market Analyst in the Photonics and Sensing Division at Yole Développement (Yole).
Pierrick works in the fields of Solid-State Lighting and Lighting Systems, carrying out technical, economic, and marketing analyses. In addition, he leads the automotive activities within the company.
Pierrick has authored several reports and custom analyses on topics such as automotive lighting, LiDAR, sensing for ADAS vehicles, and VCSELs.
Prior to Yole, Pierrick has worked in several companies where he developed his knowledge of lighting and automotive. In the past, he has mostly worked in R&D departments on LED lighting applications.
Pierrick holds a master’s degree in Electronics (ESEO — Angers, France).
This interview has been developed in collaboration with Sylvain Hallereau, Principal Technology & Cost Analyst System Plus Consulting.