Page 520 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

2�2 Benefits of the use case

1. Building the equipment knowledge base for equipment maintenance enables employees
to quickly master professional knowledge, providing more development opportunities
in the technical field, enhancing the professionalism and sense of achievement at work.
Using AI to replace operations in hazardous environments can reduce labor risks and
intensity. Accurate demand forecasting and intelligent production scheduling reduce
planning fluctuations, improve resource utilization efficiency, and enhance corporate
profitability.
2. In the "1+3+3+N" architecture, multiple intelligent platforms and AI-based applications
provide core infrastructure support for automobile manufacturing, inspiring more
innovative ideas and business models, enhancing the efficiency of the entire industry
chain collaboration, and strengthening industrial competitiveness.

2�3 Future Work

The future work of AI enabling the intelligent manufacturing of the automobile industry includes:

1. Innovative application scenarios: Optimize model development in intelligent R&D,
intelligent manufacturing, and intelligent supply chain scenarios, such as intelligent
mining tools for intelligent driving scenarios, intelligent vehicle fault diagnosis, and
integrated collaboration between production and supply.
2. High-quality dataset: Build high-quality training and evaluation datasets for scenario-
based intelligent automobile manufacturing solutions. Datasets are regularly updated
to adapt to the evolving automobile intelligent manufacturing technology and market
demand changes.
3. Continuous operation: Establish the organizational structure, resource allocation, and
operation mechanism for sustainable operation.
4. Industry standards: Develop standards covering the AI model application specifications,
data models, data dictionaries, and data security and privacy protection standards
in automobile R&D and production, laying a foundation for the healthy and orderly
development of the industry. Data models and data dictionaries standards will be
developed in ITU-T SG20, while automobile industry AI model application specifications
will be developed in ITU-T SG21.

3 Use Case Requirements

REQ-01: It is critical to build a unified system architecture which is based on one digital foundation
and intelligent platforms to support model development and multiple AI applications in various
scenarios such as intelligent vehicle R&D, production, and supply chain.

REQ-02: In production scenario, it is critical that the CV large model is deployed to improve
quality detection, efficiency and accuracy, and to reduce labour intensity and risks.

REQ-03: In R&D scenario, it is critical to use a diffusion model for text-to-image/image-to-image
generation to accelerate automobile styling concept design. Knowledge graphs, LLM and RAG
are used to improve the efficiency of vehicle design failure analysis.

REQ-04: In supply chain scenarios, it is critical to use the solver and multi-objective optimization
algorithm to implement dynamic production scheduling, shorten production scheduling time,
and reduce inventory costs.

484