Hongqi Enhances Global Luxury Vehicle Service with AI-Driven Remote Maintenance

Founded in 1953, China FAW Group (FAW) is the cradle of the automotive industry in the People's Republic of China. Its Hongqi brand, operated directly by the headquarters, is one of the oldest Chinese automotive brands. Leveraging AWS generative AI services, Hongqi Overseas' remote maintenance system optimization solution has achieved a comprehensive improvement in workflow efficiency, reducing the repair cycle from 7 days to 3-4 days. Hongqi Overseas utilizes services including Amazon Elastic Kubernetes Service (Amazon EKS), Amazon Bedrock, Amazon Nova, Amazon Glue, Amazon OpenSearch Service and Amazon Neptune.

The Automotive Industry AIGC Technology Application White Paper notes that the automotive industry is undergoing a shift from SDV (Software Defined Vehicle) to DDV (Data Defined Vehicle), with generative AI technology being the core of this technological transformation. With generative AI, automotive manufacturing can not only improve production efficiency—achieving more efficient and precise production operations and supply chain management—but also significantly enhance the consumer experience in scenarios such as smart cockpits and assisted driving. As one of the fastest-evolving sectors in China’s global expansion, the automotive sector is increasingly becoming the convergence point of "Made in China" and "Intelligent Manufacturing in China." In global competition, the Chinese automotive industry has taken the lead in achieving a leapfrog transition from mass production to intelligent innovation.

In this wave of generative AI, the question the automotive industry needs to answer has shifted from "Should we use AI?" to "How should we use AI?" Based on years of industry accumulation and experience, Hongqi Overseas has provided its own answer: "AI+Service." Harnessing the power of generative AI, Hongqi has further enhanced the intelligence of its overseas service system, embracing global customers with a luxury service experience featuring unique Hongqi characteristics.

Founded in 1953, China FAW is the cradle of China's auto industry. The Hongqi brand has actively responded to the national "Go Global" strategy, diving into the wave of global business expansion. It is deeply involved in the full process of vehicle R&D and has completed preliminary business deployments in Europe, Africa, Southeast Asia, Latin America, and the Middle East.

Currently, Hongqi covers 28 countries and regions globally with a total of 128 stores. Its performance in Europe is particularly notable: in Germany, the Hongqi E-HS9 officially became the VIP reception vehicle for Frankfurt Airport, providing VIP shuttle services for Europe's third-largest airport. In Norway, since its official launch, the E-HS9 has sold nearly 3,000 units, ranking among the top three in the large and medium-sized luxury SUV segment.

Global layout and enhancing the international influence of Chinese national brands have always been crucial parts of Hongqi's development. In the service sector, the Hongqi brand follows the same strict service standards overseas as it does domestically, collaborating with global dealers to jointly create exclusive luxury service experiences for customers and accelerate the construction of a globally leading overseas service system.

As a critical component of service, maintenance has become an important standard for testing service quality. With the iteration of Hongqi products, the quality requirements for overseas vehicle maintenance are becoming increasingly high:

• Language Barriers: When requesting support from technical experts for difficult problems, cross-regional communication is required to seek assistance from technical maintenance engineers.
• Outdated Data Collection Methods: Completed maintenance service cases could not be entered into the system as digital assets to serve as "knowledge assets" for future use.
• Complex Technical Documentation & Difficult Interaction: Existing technical materials can run into thousands of pages. Dealers or repair technicians need to manually review maintenance documents, which are often obscure, difficult to understand, and hard to interact with.

To address these issues, Hongqi Overseas decided to use generative AI as an entry point to focus on optimizing the "Vehicle Remote Maintenance Support System." In the previous project building the Hongqi Overseas EP Smart Ecological Marketing Platform, AWS performed exceptionally well and met Hongqi's overseas compliance requirements, such as GDPR. Therefore, Hongqi chose to continue its collaboration with AWS.

This collaboration with AWS focuses on the comprehensive optimization and upgrade of Hongqi Overseas' vehicle remote maintenance system:

• Building a RAG Knowledge Base to Accumulate Data Assets:
  • Amazon Glue, as a serverless data integration service, provides multiple data integration engines (Apache Spark, Ray, Python Shell) to efficiently process data such as sessions/logs, maintenance cases, and electronic manuals, implementing ETL (Extract, Transform, Load) workflows to prepare data for RAG applications.
  • Amazon OpenSearch Service supports k-NN (k-nearest neighbors) information retrieval, capable of quickly searching for the most relevant vector lists in massive data based on vectors, making it highly suitable as a vector database for RAG systems.
  • Amazon Neptune serves as a purpose-built graph database service. Amazon Neptune stores knowledge graphs generated by a separate LLM processing pipeline, where documents are transformed into graph structures before being ingested into Neptune.
• Performing RAG Retrieval to Obtain Highly Relevant Background Knowledge in Real-Time, Reducing Hallucinations, and Improving LLM Response Quality:
  • Amazon OpenSearch Service supports inverted indexes and vector retrieval, achieving efficient relevance queries across hundreds of millions of documents and improving the quality of documents read during the RAG process through multi-path recall.
  • Amazon Neptune combined with Amazon OpenSearch Service extracts vertical domain knowledge such as industry terminology. It utilizes insights into internal data relationships to resolve adjustments in cross-document complex queries, enhancing the ability to answer complex questions.
  • Amazon EKS, as a container orchestration tool, provides stable, secure, and flexible resource scheduling, simplifying operations and maintenance management complexity.

To support FAW Hongqi's overseas business, AWS provided a more intelligent solution tailored to specific and complex requirements.

As a fully managed Serverless solution, Amazon Bedrock allows Hongqi Overseas to invoke the industry-data fine-tuned Amazon Nova Lite large model service via API. Amazon Nova is a new generation of state-of-the-art foundation models from AWS, featuring advanced text, image, and video understanding capabilities, as well as leading agent and multimodal Retrieval-Augmented Generation (RAG) functions. It enables fast and cost-effective inference across intelligence categories. As a result, it supports background knowledge summarization based on knowledge base information, cross-language content understanding, and image recognition for Hongqi Overseas.

Ultimately, this provides precise diagnostic suggestions and decision support for maintenance engineers through natural language Q&A interactions.

Furthermore, Amazon Bedrock Knowledge Bases, featuring built-in session context management and source attribution, allows Hongqi Overseas complete the entire RAG workflow from ingestion to retrieval and prompt augmentation. After a maintenance work order is completed, the system uses the Large Language Model (LLM) to summarize information and conclusions from the repair process, forming a maintenance report that is uploaded to the Amazon Bedrock Knowledge Base. The Knowledge Base then uses the Amazon Bedrock Data Automation feature to automatically extract data from unstructured multimodal content—such as documents, images, audio, and video—for archiving and subsequent retrieval, thereby forming a closed loop for the end-to-end remote maintenance workflow and data flow.

Hongqi Overseas' remote maintenance system optimization solution resulted in measurable workflow efficiency improvements. The repair cycle has been shortened from the original 7 days to 3-4 days, enabling downstream maintenance engineers and dealers to achieve functions such as self-service queries, rapid response, work order tracking, and knowledge sedimentation:

• Self-Service Query: Technical experts/dealers can use AI diagnosis for real-time text/voice translation, semantic analysis, and knowledge base queries to pre-analyze the cause of faults and provide on-site technicians with repair plans and past successful cases.
• Rapid Response: On-site technicians can prepare tools and parts in advance based on expert opinions, refer to maintenance documents in the knowledge base, and begin vehicle repairs, shortening repair man-hours.
• Work Order Tracking & Knowledge Accumulation: After repairs are completed, technicians upload cases to the remote maintenance system. The system automatically summarizes the repair results, creates work order records, and submits them to the knowledge base, realizing end-to-end automation of the entire process.

When a Hongqi owner in Norway submits a maintenance work order, they are unaware of the sophisticated backend orchestration occurring behind the scenes: the system converts their voice request into technical instructions in Chinese, ultimately matching 32 similar solutions from the knowledge base. These technical intricacies transform into a tangible moment of "technological magic" for the user—accelerated response speeds, significantly shortened maintenance wait times, and improved vehicle reliability and safety. Furthermore, Hongqi users can track vehicle diagnosis and repair progress in real-time via the mobile app, enjoying a premium, efficient, and convenient service experience that reinforces their confidence and trust in the Hongqi brand.

This overseas collaboration marks a further deepening of the partnership between both parties in the field of generative AI. In the future, Hongqi Overseas will explore more possibilities of generative AI with AWS, such as building enterprise automated code review processes based on Amazon Q Developer, to further mine the potential of generative AI.

For luxury automotive brands, excellent service quality is key to winning consumers, and generative AI is an important pillar for achieving service upgrades. Empowered by intelligence, Chinese automotive technology has achieved leapfrog development, which will inevitably lead Chinese automobiles to go further globally. These Chinese vehicles, imbued with intelligent genes, are destined to reshape the rules of the century-old automotive industry.