October 24th is Xiaopeng's 1024 Technology Day. Xiaopeng released its self-developed bipedal robot , integrated land-air flying car, Fuyao architecture flagship MPV model Xiaopeng X9 and other cool technologies. As an observer of smart cars, Zhineng believes that for current car companies, the intelligent driving system is particularly critical.
Among China's new car companies, Xiaopeng Motors pays particular attention to intelligent driving. "Intelligent assisted driving for all scenarios" is also the highlight of its Technology Day. This article focuses on the new intelligent assisted driving system XNGP. It is necessary not only to explore its functional characteristics and open plans in depth, but also to understand Xiaopeng's latest technological progress in the field of intelligent driving.
Part 1. Introduction to XNGP Intelligent Assisted Driving System
● XNGP: The ultimate product form of intelligent assisted driving
XNGP is Xiaopeng Motors' latest generation of intelligent assisted driving system, which inherits and develops from the XPILOT system. Its core selling point is that it can be used nationwide and can realize full-scenario intelligent assisted driving capabilities.
The “all-scenario” goal means that intelligent assisted driving services can be provided nationwide, including:
◎ "City Navigation Assisted Driving" function covering the entire city.
◎ The “ AI Designated Driver” function provides customized services for users’ commuting and high-frequency routes .
The main selling point is AI driving, which means intelligent driving helps you drive. The main features include:
◎ Available nationwide: The AI chauffeur function will not be limited to any city, and users can set up AI chauffeur routes across the country.
◎ Learn once, and the whole process is seamless: Users only need to set up the route once, and it can be used after passing the cloud quality inspection.
◎ Continuous iteration: As user proficiency increases, the AI driver will continue to optimize routes and habits.
● XNGP system function implementation progress and driving city opening plan
◎ September 2022: Xiaopeng releases its second-generation intelligent assisted driving system XNGP.
◎ April 2023: XNGP launches urban navigation assisted driving function in Shanghai and Shenzhen.
◎ June 2023: Expanded to Beijing and Foshan urban areas, and upgraded to version 2.0 for high-speed navigation assisted driving, providing an experience close to “zero willingness to take over”.
◎ August 2023: XNGP realizes the country's first "mapless" urban navigation assisted driving, and Beijing completes a large-scale test drive experience.
◎ September 2023: XNGP announced that it will open 50 urban navigation assisted driving mainly based on "mapless" capabilities within the year.
As of now, the XNGP system has the ability to drive in cities across the country, and the opening plan is as follows:
◎ November 2023: Expanded to 25 cities, mainly covering Beijing, Suzhou, Hangzhou, Wuxi and other cities.
◎ By the end of December: Further expanded to 50 cities, including Wuhan, Chengdu, Nanjing, Xiamen and other cities, while also opening up Fujian and core cities in central and western China.
The feature we saw this time is that the Pro and Max products are upgraded, whether there is a laser radar or not , they all use a visual system as the basis. The Pro version of Xiaopeng G6/G9/P7i will be upgraded in terms of SR perception capability, high-speed NGP capability and LCC capability.
Part 2 Technical foundation: XBrain, the ultimate architecture for all-scenario intelligent driving
From a technical perspective, it mainly involves perception, planning and the subsequent support systems.
● XNet 2.0 Deep Visual Neural Network
From a practical point of view, perception is the basic skill for smart driving. The content of this year's volume is centered around vision, which is the implementation of BEV+occupying the network.
In this regard, Xiaopeng's XNet2.0 will increase the perception range by 200% and increase the types by 11. At the same time, it integrates the industry's highest-precision pure visual occupancy network, realizing the integration of dynamic BEV, static BEV and occupancy network, and is a perception architecture with the ability to understand time and space.
This is the key to the improvement of the Pro version. From the current perspective, for Chinese car companies, the role and positioning of lidar begin to revolve around redundancy and safety, and the key to improving performance is still visual capability.
● XPlanner: Neural network-based planning and control
This part is very important. Leading companies are iterating in this field . The core issue of intelligence is to understand the world in front of us and make the right response to the endless world.
◎ Long-term planning: Analyzing continuous motivations can predict your predictions.
◎ Multi-object decision-making: Based on complete surrounding information, more comprehensive decisions can be made.
◎ Strong reasoning: When you can’t see clearly or completely, you can adapt like a human.
● Key to the next generation architecture: light map, full scene, light radar
◎ Light map: Use light map technology to achieve wider functional coverage.
◎ Full scenario: Realize full scenario connection from urban/highway connection to driving/parking connection.
◎ Light radar: Gradually reduce dependence on radar and significantly reduce the overall cost of intelligent driving.
● The key to next-generation mid- and back-end capabilities: data processing efficiency
What is being compared here is the iterative ability of the technical personnel.
◎ Improved full-stack closed-loop capabilities: Full closed-loop processing efficiency increased by 150%.
◎ Full-stack simulation capability: Every code change and model upgrade is verified through end-to-end simulation, with a cumulative simulation mileage of more than 100 million kilometers.
summary
Xpeng Motors' XNGP system is currently the leader in domestic intelligent driving technology. Among the new forces, Xpeng Motors is committed to providing users with a safer, more convenient and more comfortable driving experience, leading the development direction of future intelligent driving. Of course, the volume is still very high. We always need to track the intelligent driving iteration of Huawei . This is a company that is not afraid of anyone in terms of technology.
Previous article:Who has the best autonomous driving algorithm? SAIC tied for first place in Zero Run
Next article:FAW Jiefang and Huawei comprehensively deepen cooperation, covering AI large models, autonomous driving and smart cockpits
- Popular Resources
- Popular amplifiers
- A new chapter in Great Wall Motors R&D: solid-state battery technology leads the future
- Naxin Micro provides full-scenario GaN driver IC solutions
- Interpreting Huawei’s new solid-state battery patent, will it challenge CATL in 2030?
- Are pure electric/plug-in hybrid vehicles going crazy? A Chinese company has launched the world's first -40℃ dischargeable hybrid battery that is not afraid of cold
- How much do you know about intelligent driving domain control: low-end and mid-end models are accelerating their introduction, with integrated driving and parking solutions accounting for the majority
- Foresight Launches Six Advanced Stereo Sensor Suite to Revolutionize Industrial and Automotive 3D Perception
- OPTIMA launches new ORANGETOP QH6 lithium battery to adapt to extreme temperature conditions
- Allegro MicroSystems Introduces Advanced Magnetic and Inductive Position Sensing Solutions
- TDK launches second generation 6-axis IMU for automotive safety applications
- LED chemical incompatibility test to see which chemicals LEDs can be used with
- Application of ARM9 hardware coprocessor on WinCE embedded motherboard
- What are the key points for selecting rotor flowmeter?
- LM317 high power charger circuit
- A brief analysis of Embest's application and development of embedded medical devices
- Single-phase RC protection circuit
- stm32 PVD programmable voltage monitor
- Introduction and measurement of edge trigger and level trigger of 51 single chip microcomputer
- Improved design of Linux system software shell protection technology
- What to do if the ABB robot protection device stops
- CGD and Qorvo to jointly revolutionize motor control solutions
- CGD and Qorvo to jointly revolutionize motor control solutions
- Keysight Technologies FieldFox handheld analyzer with VDI spread spectrum module to achieve millimeter wave analysis function
- Infineon's PASCO2V15 XENSIV PAS CO2 5V Sensor Now Available at Mouser for Accurate CO2 Level Measurement
- Advanced gameplay, Harting takes your PCB board connection to a new level!
- Advanced gameplay, Harting takes your PCB board connection to a new level!
- A new chapter in Great Wall Motors R&D: solid-state battery technology leads the future
- Naxin Micro provides full-scenario GaN driver IC solutions
- Interpreting Huawei’s new solid-state battery patent, will it challenge CATL in 2030?
- Are pure electric/plug-in hybrid vehicles going crazy? A Chinese company has launched the world's first -40℃ dischargeable hybrid battery that is not afraid of cold
- How to select a display detector
- Ant basic antenna array comes with PPT
- 【AT-START-F425 Review】+ ERTC and the implementation of electronic clock
- How to use FPGA coprocessor to realize algorithm acceleration
- 【Gravity:AS7341 Review】+Garbage Sorting Video
- Construction of CCS v5 software development environment for TMS320C6000
- RTOS in STM32 MCU Development
- Several Problem Solving Sharing of BQ25895
- [Zero-knowledge ESP8266 tutorial] Quick Start 15 SmartConfig one-click network configuration
- Is there anyone who uses relays to perform temperature control PID adjustment? How do you determine the PID parameters?