[Intelligent Driving Headlines] Arbe recruits well-known radar experts; Ambarella adds mass production and delivery; Catland passes grade assessment; Huawei’s car BU is rumored to be divested
To join the "Smart Car Expert Automotive Millimeter Wave Radar Industry Exchange Group" in Zhiquan , please add 18512119620 (same as WeChat), and note the company-name-position to add the group
Arbe adds award-winning radar expert Yonina Eldar from the Weizmann Institute of Science to its board of directors
Senstech deploys a variety of 4D millimeter wave forward radar products to comprehensively escort driving safety
Ambarella adds mass production and delivery! GAC Trumpchi New Energy E8 officially goes on sale!
Gatland Alps-Pro AUTOSAR project passed ASPICE CL2 level assessment
Mozhibi Intelligence: Blind Spot Warning | CAR-B15: In addition to its connotation, it also has great looks
Fului Zhixing's satellite radar solution won the honor of "Excellent Product Award" due to its many industry-leading technical features.
Muniu’s third-generation 4D imaging radar accelerates the popularization of high-speed/city-level NOA
Next-generation visual sensor company Ruisi Intelligent Chip completed hundreds of millions of yuan in Pre-B round of financing
UAE GCAA Award Honors Rohde & Schwarz for Radio Transformation at Sheikh Zayed Center
Jiang Junan, Partner and President of Xingyidao: 4D imaging radar within 0.5 degrees is expected to replace lidar
Jereh Technology’s self-developed “millimetre-wave radar video all-in-one machine” won the title of advanced technology product again
Silicon Microelectronics and Midea jointly build the "Millimeter Wave Radar Intelligent Perception Joint Laboratory"
Supporting Aanda Technology, Changdian Technology provides advanced packaging solutions for high-precision millimeter wave radar
The team of Academician Li Jun of Tsinghua University released the first autonomous driving 4D imaging millimeter wave radar data set Dual Radar for China’s urban complex climate environment
Pony.ai gets approval to start L4 autonomous truck platoon driving test in Guangzhou
Yikatong's first intelligent driving computing platform is mass-produced and delivered
WeRide was approved to launch a commercialization pilot of vehicle-mounted unmanned travel services in Beijing
Tesla FSD will enter China, can you enjoy autonomous driving for 150,000 yuan?
General Motors' (GM.US) self-driving car business Cruise plans to gradually resume services
It is reported online that Huawei’s smart car solution BU will be divested from Huawei’s system
Domestic and foreign millimeter wave radar news
| Arbe absorbs Weizmann Institute of Science award-winning radar expert Yonina Eldar joins board of directors
Arbe recently announced that Professor Yonina Eldar, an award-winning professor at the Weizmann Institute of Science, has joined ARBE’s Board of Directors. Professor Eldar is a world-recognized radar expert and will now offer her novel expertise to the company. “I am delighted to join the Board of Directors of Arbe, a company at the forefront of radar innovation poised to revolutionize safety in driver assistance solutions and autonomous vehicles,” said Professor Eldar. “Arbe’s sensing radars stand out in the industry, delivering advanced radar technology with ultra-high resolution and outstanding free-space mapping, pushing the boundaries of what the industry once thought was possible.” Eldar is a co-author of the Weizmann Institute of Signal Acquisition Modeling and Processing Laboratory (SAMPL) and Head of the Center for Biomedical Engineering.
Known for her revolutionary work in sub-Nyquist sampling, she has published over 400 journal articles, registered over 20 patents, and received over 45 awards, including the Amazon Research Award (2022), IET Radar, Sonar and Navigation Best Paper Award (2021), IEEE SPS Signal Processing Magazine Best Column Award (2021), Google Researcher Award (2020), etc. In addition, Elder is a visiting professor at MIT; a visiting researcher at Princeton University; an adjunct professor at Northeastern University; an adjunct professor at Duke University; a visiting scientist at the Broad Institute; and a distinguished visiting professor at Tsinghua University. Eldar is a member of many IEEE committees, including the IEEE Signal Processing Society Council, the Integrated Sensing and Communications Emerging Technologies Initiative, the IEEE Kiyo Tomiyasu Award Committee, and others.
She earned a bachelor's degree in physics and electrical engineering from Tel Aviv University and a PhD from MIT. "Professor Eldar brings a level of expertise to Arbe that few possess, earning her global recognition," said Arbe CEO Kobi Marenko. "We are very excited about her joining our Board of Directors." It is obvious to us the value she will bring to Arbe and, therefore, the contribution she will make to the automotive industry. "
| Senstech deploys a variety of 4D millimeter wave forward radar products to comprehensively escort driving safety
Recently, Senstech announced that it will take the lead in developing 4D millimeter-wave forward radar. Currently, two-chip cascaded 4D imaging radar STA77-6 and four-chip cascaded 4D imaging radar STA77-8 have been launched, and both products have achieved mass production. Through a large number of transmitting and receiving channels, high-resolution imaging of 4D space is achieved to meet the requirements of higher automotive automation levels.
Currently, in response to the slightly higher cost and performance improvement requirements of 4D millimeter-wave radar, Senstech is developing two 4D forward radar products using customized chips, and plans to achieve mass production in Q4 next year. The two products (8T8R forward radar and 16T16R forward radar) use customized chips and AI algorithms with built-in custom AI cores. In terms of range, speed resolution, azimuth and pitch angle accuracy and resolution, it has been significantly improved compared to previous 4D radar products.
In addition to the current 2-chip cascade, Senstech's newly developed FPGA version solution:
The solution uses a waveguide antenna, which greatly improves the radiation efficiency and can detect longer distances. The blessing of "4D millimeter wave radar + waveguide antenna" can better improve front-end performance, FOV and point cloud quality, and improve the credibility of millimeter wave radar. With the maturity of Senstech's many 4D millimeter wave radar products, the density of point clouds is getting higher and higher, and deep learning algorithms that can process more data are the future trend. With the accumulation of big data, the perception performance of millimeter wave radar will become better and better, and more comprehensive and accurate environmental perception information can be obtained, thereby improving the safety and performance of autonomous driving systems. (Photo and text source: Senstech)
| Ambarella adds mass production and delivery! GAC Trumpchi New Energy E8 officially goes on sale!
Trumpchi New Energy E8 officially goes on sale on the 17th! Ambarella Semiconductor joins hands with smart driving technology MAXIEYE to help make cars more intelligent and add new models to mass production and delivery.
As "the MPV expert who best understands Chinese people", GAC Trumpchi is a key product in its transformation into a new energy technology enterprise. Trumpchi New Energy E8 has attracted much attention since its debut. It is positioned as a medium-sized family MPV that can use fuel and electricity, and uses new energy and With advantages such as intelligence and large space, it fully meets the home needs of Chinese families. The highest price is only 229,800 yuan, and the comprehensive battery life can reach 1,200km.
(Photo source: Ambarella Semiconductor)
| Gatland Alps-Pro AUTOSAR project passed ASPICE CL2 level assessment
On November 20, the ASPICE CL2 certification ceremony for the Gatland Alps-Pro AUTOSAR project was held. After the automotive software process improvement and capability assessment ASPICE assessment, the 16 main process areas covered by Caltland’s Alps-Pro AUTOSAR project all passed the CL2 (Capability Level 2) assessment requirements with high scores, marking Caltland’s success in millimeter waves. The supporting software development and project process control of the radar chip have reached the requirements of international advanced standards.
ASPICE is a software process evaluation and improvement model for the automotive industry. It was developed by the German Automobile Manufacturers Association Quality Management Center (VDA QMC). The organization is composed of more than 600 companies in the German automotive and parts industry such as Audi, BMW, and Porsche. The ASPICE model integrates the excellent experience of the international automotive industry in software development management and is used to guide suppliers' software development process and improve the delivery quality of automotive software products. As the automotive industry continues to develop in the direction of electrification, intelligence, and connectivity, the trend of software-defined cars is gaining momentum. Currently, ASPICE has become one of the access standards for the supply chain system of global automobile manufacturers.
Different from consumer-grade and industrial-grade products, automotive-grade products need to meet more stringent safety and reliability standards. As the world's leading company in millimeter-wave radar chips, Catland has continued to strengthen the construction of its automotive chip research and development system, and has passed ISO 26262 functional safety management system certification (Level 3) and ISO/SAE 21434 automotive network security certification, integrating and building a perfect The whole life cycle development and management process of millimeter wave radar chips and put it into practice.
By passing the ASPICE CL2 international assessment this time, Gartland will rely on the support of a high-standard software development process system and adhere to the design concept of automotive-grade reliability and safety to provide customers with more transparent and reliable software and tool solutions. At the same time, this concept will be implemented into the entire process of research and development of chips, tool chains, and application solutions to ensure that more leading, safer, and more competitive millimeter wave radar chips and solutions are provided to the market. (Photo source: Gatland )
| Mozhibi Intelligence: Blind Spot Warning | CAR-B15: In addition to its connotation, it also has great looks
In September, the golden autumn of harvest, Mozhibi Intelligence relied on its strong R&D strength to launch the CAR-B15 large vehicle blind spot millimeter wave radar, ITS-CX2 visual barrier millimeter wave radar, and ITS-DX1 industrial door millimeter wave radar 3 This new product has attracted widespread attention in the industry. We will reveal the innovations and breakthroughs of these three products one by one next. The first to make a grand appearance is the CAR-B15 millimeter-wave radar warning system for large vehicle blind spots.
As we all know, most blind spot detection millimeter wave radar products are installed outside the cockpit and have to withstand wind, sun, rain and other harsh and changeable weather for a long time. Therefore, heat dissipation and waterproof performance have become important design indicators that affect the quality of the product.
Car grade waterproof interface
Integrated structural design, high vibration environment connection, waterproof and more reliable.
No welding process
Fisheye pins require no welding process, automated assembly, and higher reliability and consistency.
Automobile grade engineering plastics
Low dielectric constant and low dielectric loss, better wave transmission, 20% improvement in wave transmission performance compared to traditional wave transmission materials; excellent heat resistance, deformation temperature as high as 260 degrees.
Reliable heat dissipation design
It is made of aviation aluminum and is integrally die-cast with multiple heat dissipation points to ensure stable operation at ambient temperatures of -40~85°C.
IP67 waterproof
Excellent waterproof performance, easily meets harsh environments such as car washing, rain and snow (picture and text source: Mo Zhibi)
| Fului Zhixing's satellite radar solution won the honor of "Excellent Product Award" due to its many industry-leading technical features.
Furui Zhixing made an appearance with its satellite radar product solutions. With its excellent technical strength, it attracted many fans and attracted a large number of industry professionals to visit and exchange. The Picasso series millimeter wave radar is the industry's first millimeter wave radar based on satellite architecture. The overall solution reduces the cost of a single radar and reduces the wiring harness and installation burden. The comprehensive calculation of the overall system solution is expected to reduce installation costs by 30%.
At the same time, this is also the industry's first product to integrate radar algorithms on a domestic computing chip platform, which will greatly assist the launch of subsequent domestic assisted driving system solutions. With disruptive breakthroughs in the field of technology, Fului Zhixing's millimeter wave radar has won unanimous praise from OEMs and provided key support for the development of domestic assisted driving systems. Satellite radar also represents the centralization trend of future autonomous driving sensor standardization.
In addition, Furui Smart Radar products have also achieved innovative breakthroughs in key technical bottlenecks. This radar perfectly reuses the algorithm and transplants the existing mature radar signal processing platform algorithm to the domain controller system environment, making full use of hardware resources to improve radar performance. The current detection range capability can reach 350m, which can realize large-bandwidth communication functions and realize The high-speed transmission of radar raw data has an overall bandwidth of 6Gb. It is based on the GMSL serialization and deserialization solution and is compatible with the standard interface of existing mainstream domain control. In terms of chips, this series of products makes full use of domestic excellent chip technology to improve the cost performance of the products and provide customers with complete domestic alternatives. In addition, Picasso radar also provides the possibility to implement core high-resolution algorithms, increasing the overall imagination space of millimeter wave radar products. (Photo and text source: Furui Zhixing)
| Muniu’s third-generation 4D imaging radar accelerates the popularization of high-speed/city-level NOA
When Muniu Technology recently launched the third-generation 4D imaging radar I79, it shocked the industry.
1. Self-developed array antenna technology + real-time processing super-resolution; high target point cloud density, false target invalid point cloud rate <0.5%;
2. Vehicle-grade reliability design to meet functional safety and information security requirements;
3. It has completed 4D+ domain control integration with multiple domain control processor platforms;
4. Verified by multiple mass production evaluations, the point cloud density and target recognition capabilities have reached the best in the industry;
5. Built-in adaptive calibration and bumper matching technology reduces installation, matching and after-sales service costs.
| Next-generation visual sensor company Ruisi Intelligent Chip completed hundreds of millions of yuan in Pre-B round of financing
ALPIX-Eiger® fused high-end bionic vision sensor
According to Memes Consulting, recently, the new generation of fusion vision sensor company Ruisi Intelligent Core announced the completion of hundreds of millions of yuan in Pre-B round financing. SDIC Venture Capital and Yuanhe Chenkun jointly led the investment, and Lenovo Venture Capital, Zero2IPO Venture Capital, Guyu Jiahe, Tonge Venture Capital, China Science and Technology Advanced Industry Fund, Shenzhen Angel Fund of Funds, iFlytek Venture Capital, Zhuiyuan Venture Capital, etc. Old shareholders continue to invest. Deng Jian, founder of Ruisi Intelligent Core, said that this round of funds will be mainly used for the mass production of corporate products, accelerating the research and development of new products and the development of new fields.
Ruisi Intelligent Core is a new chip R&D and overall solution provider in the field of integrated visual sensing. The core technology is its original Hybrid Vision integrated visual sensing technology. The core product is the integrated visual sensor chip ALPIX series, which is used for smartphones, Provide integrated intelligent vision solutions in the fields of consumer electronics, intelligent security, and intelligent automobiles.
Yuanhe Chenkun said: "In recent years, there has been a scarcity of innovative technology points on mobile phones. Fusion vision sensors, as the core product of Ruisi Intelligent Core, integrate the advantages of CMOS image sensors and dynamic vision sensors, and are expected to take the lead in smart phones, consumer electronics and security Monitoring and other fields have been applied, and the growth prospects are promising. Ruisi Intelligent Core has an excellent team with comprehensive backgrounds in high-end chips, machine vision, artificial intelligence, software and hardware, etc., with strong technical and talent barriers, and is expected to grow into a benchmark enterprise in the industry. ”
Previously, Ruisi Intelligent Core has made important breakthroughs in technological innovation. Its first chip ALPIX-Pilatus based on Hybrid Vision fusion technology was released in October 2021, verifying the feasibility of the fusion vision solution. Subsequently, the first high-end integrated visual sensing chip ALPIX-Eiger was released in July 2022, which is mainly used in consumer electronics such as smartphones and action cameras.
He Zhiqiang, Senior Vice President of Lenovo Group and President of Lenovo Venture Capital Group, said: "Congratulations to Ruisi Intelligent Chip for completing a new round of financing. Visual AI has exploded into huge market potential in the fields of consumer electronics, smart security, smart cars and other fields. This is an important step in the development of AI technology." It is one of the symptoms of the coming intelligent revolution. High-end imaging technology lays the foundation for the large-scale application of visual AI. As an outstanding developer of integrated vision sensors, Ruisi Intelligent Technology has a team with experience and capabilities in algorithm, software, hardware, and chip development. Its leading integrated bionic vision technology has greatly improved the performance of event cameras and enabled visual AI to be implemented in richer scenes. It is particularly in line with Lenovo Group’s new concept of AI for all. Lenovo Venture Capital will rely on the ecological advantages of CVC 2.0 Continue to provide empowerment for Ruisi Intelligent Core and jointly promote the prosperity and development of the visual AI ecosystem.
| NXP expands S32 platform with motor control solutions for software-defined vehicle edge nodes
Specialized motor control solutions are optimized for efficiency gains in vehicle applications such as pumps, fans, sunroofs and seat positions, seat belt pretensioners, trunk openers and more.
NXP Semiconductors recently launched the S32M2, expanding its scalable S32 automotive computing platform. Specialized motor control solutions are optimized for efficiency gains in vehicle applications such as pumps, fans, sunroofs and seat positions, seat belt pretensioners, trunk openers and more. Building on the success of the S12 MagniV product portfolio, the S32M2 combines NXP’s motor control heritage with the software development strengths of the S32 platform. The highly integrated system-in-package solution adds the power and analog functions required for motor control and a rich software library to NXP's widely adopted S32K microcontrollers. It addresses the needs of the emerging software-defined electric vehicle market, enabling automakers to fully optimize product development and maximize software reuse in S32 platform implementations.
Among many types of electric motors, brushless DC (BLDC) and permanent magnet synchronous (PMSM) motors are known for their durability and robustness, small size, low weight, and high efficiency, which contribute to the energy saving and Mileage extension. The S32M2 supports simplified and more powerful motor control functions to operate efficiently while maintaining performance headroom to integrate more features and capabilities while optimizing cost. OEMs can perform motor diagnostic testing, transfer data to zone controllers and optimize for higher performance to further improve efficiency and reduce audible noise in the final application, thereby improving passenger comfort. Compared with discrete motor control solutions, the S32M2 helps OEMs shorten time to market and reduce printed circuit board size, design risk and bill of materials costs.
“By using NXP’s Model-Based Design toolkit, automakers can start Simulink simulations early in the development cycle and map the model to the S32M2 hardware. Because this is a purely software-defined approach, automakers can optimize the S32 software reuse for vehicle computing platforms while benefiting from the feature and performance enhancements based on the success of today’s S12 MagniV portfolio,” said Manuel Alves, senior vice president and general manager, General Purpose and Integrated Solutions (GPIS) at NXP Automotive Processing with the S32M2. With the addition, NXP’s S32 automotive computing platform remains consistent from top to bottom while benefiting from the vast Arm software ecosystem. NXP has a long-standing and close partnership with Arm and previous joint development, which allows NXP to combine the full potential of Arm Cortex-M processors in terms of power efficiency and safety features with new motor control solutions. combine. (Photo and text source: NXP)
| UAE GCAA Award Honors Rohde & Schwarz for Radio Transformation at Sheikh Zayed Center
Rohde & Schwarz awarded GCAA the prestigious "Radio Replacement" project, which involves upgrading more than 200 ATC radios to enhance the voice communications infrastructure already installed at the Sheikh Zayed Air Navigation Center in Abu Dhabi. The UAE's General Civil Aviation Authority (GCAA) has awarded the "Radio Replacement" project to Rohde & Schwarz. This major project aims to modernize the radio communications system at the Sheikh Zayed Air Navigation Center in Abu Dhabi. The "Radio Replacement" project involves the installation of more than 200 new radios, including major improvements to the Voice Communications System (VCS) already installed. With its strong domestic presence and comprehensive system capabilities, Rohde & Schwarz is the ideal partner for this digital transformation. Rohde & Schwarz has been present in the UAE for 30 years, demonstrating its unwavering commitment to providing cutting-edge technology and service excellence to the country. The company's market-leading next-generation air traffic control (ATC) radio, the R&S Series 5200, played a key role in winning the contract because its multiple network interfaces can be seamlessly integrated into existing infrastructure.
The signing ceremony, held at the Dubai Airshow 2023, was witnessed by Patrick Brodda, Managing Director of Rohde & Schwarz Emirates LLC, and His Excellency Saif Al Suwaidi, Director General of the GCAA. “The modernization of the radio communications infrastructure at the Sheikh Zayed Aviation Center in Abu Dhabi is one of the GCAA’s pioneering projects,” said His Excellency Saif Mohamed Al Suwaidi. “Partnering with Rohde & Schwarz will enhance the project, leveraging their extensive experience, local presence and cutting-edge technology. This collaboration will significantly aid the project development process.”
"We are honored to be commissioned by the UAE GCAA for the 'Radio Replacement' project," expressed Mr. Broda's gratitude. “This partnership reflects our commitment to delivering state-of-the-art communications solutions and our contribution to the UAE aviation industry.” Rohde & Schwarz looks forward to working closely with the GCAA to ensure the successful implementation of this project. The 'Radio Replacement' initiative is an important step in digital transformation and aims to increase the efficiency and reliability of Abu Dhabi's air navigation system, ultimately promoting the safe and seamless operation of air traffic. (Image and text source: Rohde & Schwarz )
| Jiang Junan, Partner and President of Xingyidao: 4D imaging radar within 0.5 degrees is expected to replace lidar
Recently, Jiang Jun'an, partner and president of Xingyidao, delivered a keynote speech titled "Less than 0.5 degrees: 4D millimeter wave imaging radar helps AEB and high-end intelligent driving." Jiang Junan said that compared to the core functions of lidar in SLAM (path planning) and blind filling, the role of 4D millimeter wave radar in AEB systems at L1, L2 and above may be more reassuring: it has the ability to quickly approach lidar It has excellent horizontal and pitch angle resolution and stronger environmental adaptability (working around the clock). At the same time, its cost is far lower than that of LiDAR. Its life cycle has also been verified by vehicle-level standards for a long time, and it is available in mass production. The future is bright.
It said that the company has made important breakthroughs in perception and system imaging theory. The 4D imaging radar project developed was supported by the Beijing Science and Technology Commission in 2020 and passed the acceptance this year; in addition, it completed the final defense of a key research project of the Ministry of Science and Technology earlier this month ( Next-generation ultra-high-precision 4D imaging radar). Xingyidao's products mainly serve intelligent driving of automobiles, and also cover the field of intelligent transportation.
| Jereh Technology’s self-developed “millimetre-wave radar video all-in-one machine” won the title of advanced technology product again
On November 17, the company won the 2022 "Top Ten" Advanced Technology Product Award for its independently developed "millimetre-wave radar video all-in-one machine" and was included in the "Major Innovative Technologies and Products List of Key New Energy Vehicle Enterprises in Jiangsu Province (2022)" 》.
Jereh's radar video all-in-one machine is based on a self-developed millimeter-wave radar module and integrates long-focus + short-focus dual video recognition cameras. Through real-time fusion and analysis of 2-channel video streams and 1-channel millimeter-wave radar point cloud data, it can accurately obtain detection The target's position coordinates, speed, heading angle and other operating data can be used to obtain various traffic data such as traffic flow, queue length, average speed, headway, number of stops, etc., which can detect traffic events and output alarms. The product is designed for urban roads, highways, intelligent network and other scenarios. The detection area covers 300 meters and 10 lanes, achieving all-weather, high-precision, multi-target real-time fusion perception analysis of people, vehicles, roads and environment information, serving urban traffic management, Business applications such as signal control, traffic guidance, vehicle-road collaborative roadside traffic status awareness, etc. (Photo and text source: Jerry Technology)
| Silicon Microelectronics and Midea jointly build the "Millimeter Wave Radar Intelligent Perception Joint Laboratory"
Recently, Foshan Midea Global Innovation Center, Midea Central Research Institute, through the Perception Technology Subcommittee, jointly established various business units and Silicon Microelectronics to establish the "Millimeter Wave Radar Intelligent Perception Joint Laboratory" to help realize the perception layer technology of smart home appliances and smart homes. Smarter non-contact sensing and interactive functions The joint laboratory established by both parties aims to leverage Midea Group’s deep roots in the home appliances industry, combined with SiDian Micro’s millimeter wave radar technology and easy-to-use chip products, to accelerate millimeter wave radar Promote sensing technology and application technology in the smart home market, and jointly explore the sustainable and organic development of the intelligent application ecosystem.
As early as 2021, SiDian Micro Laboratory has passed the authoritative SGS ISO17025 laboratory quality system certification. R&D engineers can apply SiDian micro-millimeter wave sensor chips, reference designs and other products and technologies to project research and development here, so that the end products have the characteristics of high integration, easy development, accurate identification and anti-interference, etc., and greatly reduce the development time.
The "Millimeter Wave Radar Intelligent Perception Joint Laboratory" located in Midea's Global Innovation Center allows Midea Group's R&D engineers to use SiDian Micro's experience and technical support in chip performance, antenna performance, domestic and foreign standard certification, environmental reliability testing, etc. Design intelligent terminals with high reliability and leading technology that comply with domestic and foreign standards. The two parties will steadily enhance their R&D experience and project cooperation in many millimeter wave radar application fields through common technology research, testing and certification. Promote the upgrade of millimeter wave radar sensor testing and application capabilities within the industry.
ICL1122 SoC is an innovative μA-level 24G single-transmit and dual-receive millimeter-wave sensor chip from Silicon Microelectronics, which combines ultra-low power consumption (70μA) with super detection capabilities for wide-range precise position sensing. The chip adopts a fully integrated CMOS architecture, integrates a complete millimeter wave sensor system and signal processing, and has unlimited cascading functions. It has few peripheral components and is simple and easy to use. It can meet the ultra-low power consumption requirements of battery-powered application scenarios and the ultra-wide range detection capability requirements of outdoor scenarios. Suitable for multi-person trajectory detection and vehicle and pedestrian monitoring in rich application scenarios. (Picture and text source: Silicon MicroLabs )
| 99.03%! Professor Zhao Yaqin’s team from Harbin Institute of Technology proposed a new high-precision MIMO millimeter wave radar micro-gesture recognition method
Recently, Professor Zhao Yaqin of Harbin Institute of Technology and others published the latest article "High-precision gesture recognition based on DenseNet and convolutional attention module" in the Journal of Electronics and Information.
Author: Non-contact gesture recognition is a novel interaction method that can be widely used in fields such as augmented reality, virtual reality, smart home and smart medical care. It has become a hot research topic because we can use it to detect small gestures. This article introduces a new method using millimeter wave radar for small gesture recognition.
We used four radar panels to obtain the reflected signals of gestures. We then analyze these signals to detect gestures and hand positions. Through a series of data processing steps, we are able to gain a more comprehensive understanding of the motion characteristics of gestures. Finally, we used a special neural network to recognize 12 small gestures. This network was very accurate, with an accuracy of 99.03%. The network can also focus more specifically on the first half of a gesture, improving gesture recognition accuracy.
A new micro-gesture recognition method based on MIMO millimeter wave radar is proposed, which improves the gesture recognition effect and achieves high-precision gesture recognition. On the basis of existing research results, the MMWCAS radar composed of 4 AWR1243 radar boards cascaded was used to collect gesture echoes, and a gesture recognition data set containing 12 gestures and 600 samples for each gesture was constructed to provide gesture data processing. The work provides strong data support. Then, the range-Doppler spectrum and 3D point cloud were used for target detection, and the range-time spectrum, Doppler-time spectrum, micro-Doppler-time spectrum, azimuth-time spectrum and The five characteristics of the pitch angle-time spectrum are more comprehensive than other studies and can accurately characterize a variety of micro-movement gestures. Finally, a gesture recognition network based on DenseNet and CBAM is proposed, using the mixed feature map after data expansion for gesture classification. Experiments show that the network focuses on the first half of the gesture action and can be flexibly adjusted, reaching 99.03% The recognition rate achieves high-precision gesture recognition.
| Help Anda Technology, Changdian Technology’s advanced packaging solutions for high-precision millimeter wave radar
In the high-precision millimeter-wave radar market, Changdian Technology cooperates with many leading millimeter-wave radar chip customers at home and abroad to develop, using packaging technologies such as eWLB and FC flip-chip to meet the needs of customer products with multiple transceiver channels, integrated antennas, and low power. consumption needs and help customers enrich their product lines. At present, Changdian Technology has achieved large-scale mass production of millimeter wave radar products.
The hardware core of millimeter wave radar includes MMIC and antenna. The system architecture consists of antenna, transceiver module, power management, signal processing and other parts. Chip packaging methods mostly use flip-chip packaging, fan-out packaging, and AiP (Antenna in Package) packaging that integrates antenna and radar transceiver chips.
Aanda Technology has maintained a long-term cooperation with Changdian Technology to provide the market with millimeter wave radar solutions that are easy to integrate, have low energy consumption and are highly cost-effective. Based on Antenna Technology's CMOS millimeter wave radar SoC chip, the newly launched 60GHz AiP (Antenna in Package) and 77GHz AiP series of millimeter wave radar chips further reduce the complexity and cost of millimeter wave radar systems. During the cooperation process, Changdian Technology has provided complete flip-chip (FCCSP) integrated antenna AiP and other packaging products in view of the multi-scenario application characteristics of the product, achieving small size, high yield, high reliability and low cost of the product, which can meet the needs of Application needs in multiple fields such as automotive intelligent driving, intelligent cockpits, smart homes, drones, and intelligent transportation. (Photo and text source: Changdian Technology )
| Waveguide antenna helps enhance the performance of automotive millimeter wave radar
The use of waveguide antennas can improve the performance of millimeter wave radar. During the transmission of electromagnetic waves, waveguide antennas can bind them in hollow metal, greatly reducing energy loss. Therefore, waveguide antennas have low insertion loss, enhanced shielding and high-power processing capabilities, making it easier to implement large-bandwidth operating modes and obtain higher resolutions, which can enable millimeter-wave radar to play a role in automobile collision avoidance, pedestrian detection, traffic monitoring and other fields. greater effect.
There are currently two mainstream varieties of vehicle-mounted waveguide antennas: slot antennas and cavity antennas. Compared with traditional waveguide antennas, they are miniaturized, have good heat treatment capabilities, and have no inferior performance, so they can be mass-produced and put into vehicles. A slot antenna, also known as a slot antenna, consists of one or more apertures created by removing material from a conductive surface. It has the advantages of wide bandwidth, high efficiency and strong radiation pattern control. It can easily support the 76 to 81GHz frequency band, with more accurate measurement and more flexible frequency allocation. Regardless of the gain of the antenna element, the losses are negligible. The radiation pattern is very wide and stable, making it ideal for short-range radar applications. In addition, if there are assembly defects in the antenna, the slot antenna can still limit the field within the waveguide channel, improving fault tolerance.
The waveguide cavity antenna is an antenna array deployed in three directions of xyz and has more obvious advantages in multi-chip cascaded 4D millimeter wave radar. With a larger number of transmitting and receiving antennas in the radar system, the design and feeder routing will become more complex. Since the waveguide cavity antenna can be deployed in three directions, the design becomes more compact, reducing the size of the antenna and radar, while also increasing the radar detection range.
In the future, research on the plastic metallization process of 3D waveguide antennas will become one of the development ideas. Antennas using this process are small in size, have extremely low insertion loss, and are highly reliable. They can provide higher gain for millimeter wave signal return and achieve Higher precision object detection function. At present, HUBER+SUHNER has completed mass production, but domestic companies are generally troubled by problems such as high costs and low qualification rates, and they still need to work hard to overcome technical difficulties.
| The team of Academician Li Jun of Tsinghua University released the first autonomous driving 4D imaging millimeter wave radar data set Dual Radar for China’s urban complex climate environment
Recently, the team of Academician Li Jun of the School of Vehicles and Transportation of Tsinghua University released Dual Radar, the world's first autonomous driving 4D imaging millimeter wave radar data set for China's urban complex climate environment, and made it free and open source, which provides a basis for research on 4D millimeter wave perception algorithms in complex climates. A new test benchmark is provided. This dataset collects two mainstream types of 4D radar simultaneously for the first time and can be used to explore perception algorithms under different radar point cloud densities and noise levels.
Dual Radar data set collection vehicle
The Dual Radar data collection vehicle is equipped with cameras, lidar and two mainstream 4D millimeter wave radar sensors: ARS548 RDI from German mainland and Arbe Phoenix from Israel. It can operate in various climates (sunny, cloudy, rainy, etc.) in the complex traffic environment of domestic cities. Images, laser point clouds and two types of 4D millimeter wave point cloud multi-modal data are collected simultaneously under different lighting conditions (strong light, low light, backlight, etc.). The total collection mileage exceeds 600 kilometers. The annotated data consists of 151 continuous sequences, each sequence lasts 20 seconds, contains 10,007 synchronized and annotated key frames, and 103,272 three-dimensional targets are annotated. The data set can be used for research on autonomous driving perception algorithms, including three-dimensional target detection and tracking in complex scenes, and can be used to explore the application of 4D millimeter wave radar in autonomous driving perception.
Academician Li Jun's team is one of the first teams at home and abroad to conduct research on autonomous driving perception algorithms in 4D millimeter waves. It has made progress in the fusion of cameras and 4D millimeter waves, and the fusion of lasers and 4D millimeter waves. A multi-modal fusion algorithm VVE-Fusion is proposed to thicken radar point cloud information based on image context texture features. It uses image semantic segmentation information to increase millimeter wave point cloud density and filter foreground information, effectively improving detection accuracy and solving 4D millimeter wave radar problems. The point cloud is too sparse and the detection accuracy is low;
A multi-scale fusion 3D target detection algorithm M2-Fusion based on cross-modal interaction of 4D millimeter waves and laser point clouds is proposed. The attention mechanism is used to realize shallow and deep information interaction between the two modalities, and near the regression target key points. Perform multi-scale feature extraction to achieve a balance between accuracy and efficiency, effectively improve the accuracy of three-dimensional dynamic target detection, and enhance the long-distance detection capability of lidar. Relevant results were published in the field journal T-VT and the conference IROS.
Typical scenes and data visualization of Dual Radar data set (rainy day, night and tunnel scenes, etc.)
Academician Li Jun's team has rich technical accumulation and achievement output in the field of autonomous driving. It has established a multi-sensor interpretable fusion framework for intelligent driving perception safety, and proposed a camera-lidar fusion high-end based on "appearance-motion" combination optimization. The reliable 3D multi-target tracking algorithm CAMO-MOT ranks first in the nuScenes 3D tracking evaluation list of the famous large-scale data set in the field of autonomous driving (for 7 months), and builds an intelligent driving fusion sensing invention patent group (U.S. authorized invention patent 10 items, 50 invention patents authorized by China), the team’s research results have been published in T-ITS, T-NNLS, T-VT, T-IV, Information Fusion, Pattern Recognition, CVPR, ICCV, ICRA, IROS, ITSC and other automatic Driving in prestigious journals and international academic conferences. (Photo and text source: School of Vehicles and Transportation, Tsinghua University )
| Pony.ai gets approval to start L4 autonomous truck platoon driving test in Guangzhou
Recently, Pony.ai announced that it has obtained the first self-driving truck platoon driving test license in Guangzhou, and is allowed to conduct L4 platooning self-driving tests on Guangzhou’s open roads. This provides strong support for Pony.ai to further develop large-scale commodity fleet autonomous driving transportation business.
Pony.ai's self-driving truck fleet is a "1+N" form, that is, one self-driving truck leads multiple L4 self-driving trucks to drive safely on the highway. A total of five self-driving trucks have obtained platoon driving test licenses this time. This "1+N" formation of autonomous driving greatly reduces operating costs and improves transportation efficiency. In addition, by eliminating the difference in braking time between front and rear vehicles and shortening the following distance, the vehicle in front can break the wind for the vehicle behind, effectively reducing air resistance, thereby reducing carbon emissions, achieving green travel, while improving road usage efficiency and alleviating road congestion. , reduce traffic accidents.
In December 2020, Pony.ai obtained the first self-driving truck test license in Guangdong Province. On the basis of this test license, Pony.ai continues to polish and improve in technology research and development, system optimization and actual testing, fully meeting a series of requirements such as autonomous driving capabilities, test plans and content required to obtain this formation driving test license. , thus successfully obtaining the formation driving test license. This not only demonstrates the recognition of Pony.ai’s technical strength by the Guangzhou municipal government departments and tripartite institutions, but also injects new impetus into Pony.ai’s future development.
At present, Pony.ai has carried out self-driving truck trunk logistics and dedicated line logistics and freight exploration in many places in China, realizing autonomous driving of trucks under different roads, weather and other conditions. In June 2022, Pony.ai’s self-driving trucks were recognized in the Guangzhou mixed-driving demonstration operation model catalog and obtained demonstration operation qualifications; in September 2023, Pony.ai was awarded the intelligent connected heavy-duty truck in the Beijing Intelligent Connected Vehicle Policy Pioneer Zone Demonstration application notice, its self-driving trucks will work with Sinotrans to provide demonstration freight services in real scenarios.
As of now, Pony.ai’s self-driving trucks have a test mileage of more than 3 million kilometers, a commercial operation mileage of more than 610,000 kilometers, and a total cargo weight of nearly 40,000 tons. In the future, Pony.ai will continue to deepen technology research and development and application implementation to help the logistics industry achieve intelligence and modernization. (Photo and text source: Pony.ai )
| Yikatong's first intelligent driving computing platform is mass-produced and delivered
Demonstration of the parking function of Yikatong·Tianqiong Pro intelligent driving computing platform. Data map
At the beginning of this month, Yikatong Technology, which has been deeply involved in Wuhan Economic Development Zone for seven years, announced that its first intelligent driving computing platform - Yikatong Tianqiong Pro parking and parking integrated intelligent driving platform has officially achieved large-scale production on the Lynk & Co 08 model. Product delivery. Cai Lili, senior vice president of sales and marketing communications in China, said that the intelligent driving platform was officially mass-produced and delivered to cars, which is an important turning point for Yikatong Technology. It means that Yikatong Technology has officially The complete capabilities of “smart cabin + smart driving”.
According to reports, the "Huashan No. 2" A1000 intelligent driving chip equipped on the platform comes from Black Sesame Intelligent Technology Co., Ltd. (hereinafter referred to as "Black Sesame Intelligence") located in Qingshan District, Wuhan. This intelligent driving computing platform is produced at Wuhan Dongjun Automotive Electronics Co., Ltd. (hereinafter referred to as "Dongjun Auto") located in Wuhan Economic Development Zone. Six years ago, as early as the early days of Ekatong's establishment, it reached a cooperation with the "old" auto parts company Dongjun Industrial. The two jointly established Dongjun Automobile and jointly entered a new track for the next generation of automobiles.
It is understood that Yikatong Technology is currently focusing on the continuous layout of the intelligent connected automobile industry chain, and is working with many companies from Wuhan to establish an industrial ecosystem and jointly enter a new track for the next generation of automobiles. What is an intelligent driving computing platform? Cai Lili told reporters: "If the smart cockpit is compared to a Chinese essay question, it focuses on the emotional and emotional experience of thousands of people; smart driving is more like a mathematical application question, with different calculation methods, but the answer is the same. In terms of emotional intelligence and The computing platform can be regarded as the 'central brain' for intelligent driving. "The intelligent driving computing platform is responsible for connecting various sensors on the car body, such as lidar, Cameras, etc., and provide analysis and decision-making through its own computing power, which is the core key to realizing intelligent driving. ” ( Picture and text source: Yikatong Technology )
| WeRide was approved to launch a commercialization pilot of vehicle-mounted unmanned travel services in Beijing
On November 17, WeRide obtained the pilot notice for the commercialization of "unmanned in the car and remote outside the car" travel services for passenger cars in the Beijing Intelligent Connected Vehicle Policy Pilot Zone (referred to as the "Pioneer Zone"), and was approved to operate in Yizhuang, Beijing. Carry out unmanned self-driving travel service (Robataxi) charging in the car.
Since it was first approved to carry out unmanned road test business in the demonstration area in October 2022, WeRide has gradually been approved to carry out the "unmanned vehicle off-road remote stage" demonstration application business and the airport expressway test business in the demonstration area, and now it has won the first place The district’s pilot notice for the commercialization of unmanned autonomous driving travel services in vehicles has been among the fastest in the industry.
Citizens in Beijing experience WeRide’s self-driving travel service
After the service is open to the public, citizens can call WeRide's unmanned self-driving travel service vehicles within the designated area through the WeRide Go App and view the estimated order cost in real time. After the vehicle arrives at the designated boarding point, passengers can scan the code to verify their identity and enjoy the easy and convenient travel services brought by autonomous driving. The service scope of WeRide's fully autonomous self-driving travel service vehicles covers popular destinations such as core subway stations in the Beijing Economic and Technological Development Zone, residential areas, key business districts, large office parks, advanced manufacturing enterprises, etc., with a total of 242 pick-up and drop-off stations, supporting 1-3 passengers can experience it together. The service time is from 9:00 am to 17:00 pm. (Source: Sohu Auto)
| Tesla FSD will enter China, can you enjoy autonomous driving for 150,000 yuan?
"Tesla FSD is ready and will soon enter the Chinese market!" Twitter blogger @WuWa broke the news.
Another Twitter blogger @Teslascope said that Tesla is already providing user manuals containing FSD beta versions to European, Chinese, and Australian car owners. It seems that, as the core competitiveness of Tesla cars, FSD (Full Self-Drive, fully autonomous driving) is really going to enter the Chinese market.
The news that FSD is about to enter the Chinese market was exposed at a coincidental time. On the opening day of the Guangzhou Auto Show on November 17, the Ministry of Industry and Information Technology, the Ministry of Public Security, the Ministry of Housing and Urban-Rural Development, and the Ministry of Transport jointly issued the "About the Development of Intelligent Network Connectivity". Notice on the Pilot Work for Car Access and Road Access" (hereinafter referred to as the "Notice").
The main content of the "Notice" is to propose more complete access rules for smart cars and L3/L4 autonomous driving, and it is determined that some products that support autonomous driving will be selected for pilot testing. The "Notice" lays the legal and regulatory basis for the popularization of autonomous driving in China, which is a decisive factor for FSD to enter the Chinese market.
Secondly, although Tesla does not pay attention to marketing, it does not need marketing. Tesla’s popularity mainly comes from Musk, the world’s largest Internet celebrity. However, in the Chinese market, it is not just Musk’s name that can make consumers obediently pay for Tesla. China is about to select some product pilots that support autonomous driving. If Tesla can become the first car company to participate in the pilot, its popularity and reliability in the eyes of consumers will naturally increase significantly.
Based on a pure visual solution, Tesla can decentralize FSD to models worth more than 100,000 yuan. According to Musk's plan, Tesla's first low-priced model may be named Model 2. It is expected to be put into mass production in 2025. It will use an integrated die-casting platform and can be considered a scaled-down Model Y. The price will be about 150,000 yuan. Get started. Domestic vehicles that support high-end assisted driving or have hardware that has the potential to upgrade to autonomous driving are usually more expensive. The entry of low-priced Tesla autonomous driving will inevitably have a serious impact on the domestic automobile industry.
The arrival of FSD will have a certain impact on domestic car companies, but domestic car companies do not need to panic. Even from a technical and cost perspective, Tesla may not have an advantage over domestic car companies. The ranking of self-driving companies released by the research organization Guidehouse Insights at the beginning of this year showed that among the 16 companies surveyed by the organization, Tesla ranked first from the bottom. Domestic companies such as Baidu and Pony.ai were all ranked higher than Tesla. Data from Guidehouse Insights shows that there are many domestic companies with self-driving technology superior to Tesla. Even if the car companies’ own technology is lagging behind, they can still choose to use Baidu and Pony.ai’s technology. (Photo and text source: Lei Technology)
| General Motors' (GM.US) self-driving car business Cruise plans to gradually resume services
Zhitong Finance APP has learned that Cruise, the self-driving car business of General Motors (GM.US), plans to gradually resume services. It follows a series of missteps that led the company to halt all self-driving taxi operations last month. Cruise President Mo Elshenawy told employees in an email on Wednesday that the company would eventually restart operations in one city. The company will focus on revamping Chevrolet Bolts in the near term while delaying plans for its self-driving Origin cars, according to an email confirmed by a spokesperson.
"Once we take steps to improve our safety culture and rebuild trust, our strategy is to relaunch in one city and prove our performance there before expanding," Cruise said in an emailed statement. The decision marks the A shift in strategy for the company, which now operates in multiple cities including San Francisco, Phoenix and Austin, and until recently had plans to expand to other cities.
But last month, California's Department of Motor Vehicles suspended Cruise's license to operate self-driving cars in the state, accusing the company of withholding key video of an accident involving a pedestrian. Shortly thereafter, General Motors suspended its entire fleet, and in the past week, Cruise CEO Kyle Vogt abruptly resigned. Following Vogt's departure, Elshenawy was named president and chief technology officer, but he did not specify in which cities the company would resume service. In related news, the company's options may be Texas or Arizona.
| Internet transmission Huawei’s smart car solution BU will be stripped from Huawei’s system
On November 23, it was reported that
Huawei's smart car solution BU (hereinafter referred to as "Car BU") will be separated from Huawei's system
, with an overall valuation of 250 billion yuan. After the change of hands, the largest shareholder is Chongqing State-owned Assets Supervision and Administration Commission. ; Changan is the only car company participating in the acquisition, acquiring a 15% stake in Che BU for a price of 37.5 billion, and the Bingzhuang Group holds about 5%. There is also news that Changan Automobile will acquire 30% of the shares of Auto BU at a price of 300 billion yuan.
-
It can still provide parts and components to Huawei Smart Car as a core supplier;
-
And it can take advantage of Chongqing’s geographical advantages to deepen cooperation with car companies such as Changan and Thalys;
-
The new entity also means that Auto BU is expected to lift the blockade of some supply chains and obtain some important chip resources to provide more competitive products.
Currently, Huawei's cooperation with car companies is divided into three categories: vertical parts supply, HI (HuaweiInside) model and Huawei Smart Selection model. The HI model provides car companies with full-stack smart car solutions. The Huawei Smart Selection model is provided by Huawei In-depth participation in product definition, styling design, marketing, user experience and other aspects.
Smart Car Experts focuses on the smart sharing and interactive communication platform for the smart connected automobile industry. It has established 10+ industry communication groups and regularly broadcasts online special live broadcasts.
We sincerely invite you to join the Smart Car Expert millimeter wave radar industry WeChat exchange group :
The group includes heads of 500+ autonomous driving companies in China, module engineers of OEMs, general managers of companies, R&D directors, chief engineers, professors and experts from universities, etc., gathering industry elites from related industry chains. Welcome to join the group to communicate.
Please add Xiaoyi@zhichangjia WeChat : 18512119620 to apply to join the group.
Autonomous Driving | LiDAR
Millimeter wave radar | Automotive electronics
1500+ professional communities invite you to join!
Reply "Join group" in the background to join
If you think it looks good, click "Looking"!