Artificial intelligence is leaping from the academic field to a new stage of practical application, and large model-driven swarm intelligence technology is becoming the core driving force to promote innovation.
For the automobile industry, the boarding of large models is still in its infancy.
As of June, more than 20 domestic car companies have implemented large models, and the scene mainly covers intelligent cockpit and intelligent driving.
in addition, a few large models are used in design, production and sales.
The whole industry is faced with providing users with more personalized and cost-controllable service solutions more efficiently through the integration of man and machine in the full-contact driving of the whole user.
According to Feng Jingfeng, CEO of Lutes, Lutes has a smart driving team of less than 400.
compared with 2000 or even 3000 people in other car companies, the win-win situation of “minimizing personnel and maximizing the effect” is due to intelligent systems.
Thus it can be seen that the overall cost leadership has undoubtedly become the basis of the competition of automobile enterprises, and the intelligence of artificial intelligence will become the winner of the decisive battle.
Therefore, he Sicong, vice president of Ellabee, said at the automobile blue book forum that the most simple application of artificial intelligence in automotive software is a supplement to the human service knowledge base for automobile diagnosis.
as there will be more and more problems in the electronics of our whole vehicle, the biggest application of artificial intelligence for after-sales service engineering will be to reduce cost and increase efficiency.
At present, it seems that how to integrate traditional diagnostics with more intelligent new technologies is on the eve of the outbreak.
With the increasing maturity and popularization of OTA+ technology, diagnosis and analysis technology is playing a more and more important role in the automobile industry.
From remote diagnosis and machine diagnosis to flexible data acquisition, big data visualization and intelligent driving log collection and analysis, these technologies not only promote the development of automobile intelligence, but also promote the host factory to make a profound transformation in the traditional after-sales and service field.
Under the background of the rapid iteration of technical means, the mainframe factory is facing unprecedented opportunities and challenges.
The traditional after-sales and service models are gradually replaced by new service models such as data early warning, active service and fault transmission chain truncation.
The new mode relies on the in-depth grasp of data and signals, as well as the accurate correlation between failure modes and data information, so as to achieve real-time monitoring and predictive maintenance of vehicle status, which means a qualitative jump in the initiative and automation of diagnosis technology.
However, it is difficult for traditional dealers and maintenance personnel to have a deep understanding of complex technologies such as data, signals, logic, strategies and so on.
How to quickly integrate the intelligent maintenance and diagnosis technical solutions into the current vehicle maintenance and diagnosis practice and promote cost reduction and efficiency by intelligent means is the focus of the whole industry.
Huawei released the VHR cloud service 3.
0 cloud magpie model, and intelligent car diagnosis has been put into public view.
It is reported that the large model of Qianzhu Yunque supports question-and-answer interaction, enter fault description, and the large model can automatically pass semantic analysis, understand the problem, carry out intelligent triage, make diagnosis plan, and generate diagnosis conclusions and repair suggestions.
The automatic execution of the whole process further reduces the original hour-level diagnosis time to the minute level, which is significantly improved compared with the existing remote diagnosis capability.
Is AI a strong assist or not? Under the background of the new competition of changing style once a year, cars are no longer small or big changes in the traditional sense.
At the same time of high-frequency iteration of different brands and different models of new energy vehicles, many software versions and complex data structures have become one of the challenges of intelligent diagnosis.
The key word for Huawei’s solution is undoubtedly AI.
On the other hand, the mainframe factory, in building and managing its own intelligent platform, uses artificial intelligence to improve its own system efficiency, and it is not simple to make the best use of the value of AI in the field of diagnosis.
In the past six months, our team has flipped through a large number of AI papers and tried different large models at home and abroad to analyze quality and after-sales problems, but the results are not satisfactory and the after-sales department does not pay the bill.
The technician of a mainframe factory discussed with Ellabee, and this exclamation is no longer an exception.
As a matter of fact, there is an obvious deficiency in the logic deduction ability of AI large model at present, and it is unable to give accurate information in the fields with professional requirements, such as failure analysis and maintenance scheme.
Thus it can be seen that the ability of underlying data and knowledge base directly affect the intelligent effect of AI in diagnosis.
When discussing the field of R & D quality and after-sales technology, Chen Zixin of Ellabee Diagnostic BU elaborated in detail the key to how AI and computer accurately understand and analyze vehicle logic and signals.
He pointed out that the core of success does not lie in the AI technology itself, but depends on three key technical preconditions: 1.
The perfection of the knowledge graph: after more than ten years of development, the knowledge graph has made great progress, and its application in the field of diagnosis can better understand and analyze the correlation and dependence between the various parts of the automobile, and integrate the information of various parts, fault types and solutions.
Form a comprehensive knowledge base.
It should be emphasized that the construction of a knowledge graph suitable for the automotive industry needs to be combined with rich project experience and continuously iterated and maintained by engineers with experience in vehicle subsystem development and operation.
On this basis, combined with the AI model, through the strong combination of knowledge graph and AI, the development efficiency of knowledge graph can be improved, and the diagnosis time can be shortened from hour level to minute level.
What can not be ignored is that the difficulty and cost of constructing knowledge graph should not be underestimated. 2. The integration ability of the platform: the complicated and independent systems of the mainframe factory also pose a challenge to the diagnosis.
The intelligent diagnosis platform needs the necessary strong integration ability to dock and integrate dozens of system data.
In-depth understanding of the mainframe factory business scenarios and underlying design.
Professional know-how in diagnostics, network messages, MPU logs, OTA, accessories, maintenance materials, etc.
, and know how to combine these data to assist AI for efficient analysis. , 3. Fine implementation of data governance: in the process of comprehensive integration and management of source data, special attention is paid to the detailed carding of key information sources such as SSTS (Service support system), DEFMA (Design failure Mode and impact Analysis), maintenance manuals, engineering schematics and so on.
Through the advanced atomization splitting technology, the complex data sets can be refined to the most basic components, which is closely related to the vehicle master data management.
The dense linkage mechanism ensures that each model data strictly follows the vehicle-specific configuration properties and fine breakpoint information.
In addition, in order to maximize the value of data, it is necessary to deeply integrate diagnosis with automobile lifecycle software management platform, and closely associate problem analysis with high-frequency OTA activities, so as to improve data value.
The foundation is not solid, even if the mainframe factory invests a lot of resources in AI development, it often gets twice the result with half the effort.
Chen Zixin stressed that the intelligent management and application of vehicle data is not achieved overnight, but a long-term and continuous process, which needs to be optimized and updated according to dynamic vehicle information, R & D information and after-sales information.
Therefore, a long-term and stable partner in the automotive industry will be an indispensable and powerful assistant for the mainframe factory in the field of AI application.
Whether the three can break away from the lower volume price, the up-roll technology is the healthy competition.
Intelligent diagnosis tuyere has come, AI manufacturers flock to, but from the industry practice, the Internet and AI manufacturers, which are not deeply integrated with the automobile industry, will inevitably face disobedience.
This is the reason why the mainframe factory and AI are not holding hands quickly.
Alabi believes that intelligent diagnosis is currently a technical shortcut to reduce cost and increase efficiency.
But to do a good job of intelligent diagnosis, we need to ask “three can”, that is, whether the platform can be integrated, whether the intelligence can be landed, and whether the data can be based strongly.
With the complexity of the automobile system, the traditional decentralized platform has been difficult to meet the needs of efficient and accurate data processing.
Through the integration of TCE and BOM, Alabi’s A6 platform can seamlessly dock dozens of system platforms of the car factory, break the data isolated island and realize the interconnection of data.
At the same time, A6 can collect, organize and analyze data from different data sources, and realize the standardization, unification and visualization of data.
It provides a valuable data governance and fusion base for the construction of intelligent diagnosis platform.
Although the knowledge graph is becoming more and more mature in other industries, the application in the field of automobile diagnosis is just beginning.
Unlike most of the AI manufacturers that stay in POC and PPT, the Erabbi knowledge graph development team deeply ploughed intelligent diagnosis and accepted the map landing solutions of many manufacturers.
In view of the special attribute of long cooperation cycle of intelligent diagnosis project, stability is particularly important in the period of industry excitement.
As an industry leader in the field of OTA+, Ellabee has not only long-term low-level development experience in after-sales diagnosis, remote diagnosis, data acquisition, big data platform and OTA development, but also long-term operation experience in OTA and after-sales diagnosis.
Ellabee deeply integrates these capabilities with AI to create a new generation of intelligent diagnosis platform.
The platform not only solves the problem of high difficulty and high cost in the construction of knowledge graph, but also greatly shortens the management and editing time associated with thousands of signal screening and failure model.
In terms of question retrieval and question answering, we take advantage of the large language model to provide a more accurate and efficient solution.
In practical application, this intelligent diagnosis platform can not only effectively maintain the logic of signal, data, model and diagnosis strategy, but also connect each model closely through the knowledge graph, and deepen the function of intelligent service to every link.
This comprehensive intelligent upgrade has brought unprecedented convenience and efficiency for vehicle maintenance and troubleshooting.
Of course, it is not enough to have a system in the process of intelligence.
in order to better provide data enabling services, more and more host factories begin to adjust their organizational structure, and even set up special intelligent modeling and operation teams.
to further improve the efficiency and capability of integrated intelligent operations.
Diagnosis is neither the goal nor the end, the car is to achieve one-stop vehicle health management.
As Wan Gang said, the next step in car intelligence is to be safer and reduce accidents, which is the most important for intelligence.
Alabi integrates OTA, intelligent diagnosis and automotive software management platform, focusing on software management, diagnosis, repair and updating to establish an automotive health system in order to improve quality and efficiency, reduce cost and increase efficiency.
, return to the first electric network home page >.