Connecting the world and beyond

Vertical Industry & Knowledge Engineering

9.0.1.jpg

This category aims to bridge the "last mile" of deploying AI technologies and foundation models into various vertical business scenarios. It standardizes domain-specific knowledge engineering graph extraction, industry-specific foundation model fine-tuning construction, and specific implementation functions and evaluation metrics of AI in telecom smart customer service, home service robots, industrial inspection, power maintenance, and other specific industries.

9.1 Knowledge Engineering & Fine-tuning

Specifically regulates dataset construction evaluation standards required for customizing foundation models for vertical businesses, as well as enterprise-level unstructured knowledge fusion, graph extraction, and automatic reasoning application system frameworks.

ITU-T F.TAKEA: Framework and technical requirements for knowledge engineering applications based on foundation model (2025)
Standardizes the technical framework requirements for implementing enterprise-level knowledge engineering applications combining foundation model advantages, including cross-modal knowledge extraction and knowledge graph chain fusion reasoning handling complex logic via AI.

ITU-T F.CETD: Requirements for the Construction and Evaluation of Fine-Tuning Datasets for the Training of Domain-specific Models (2025)
Targeting the custom development of proprietary foundation models for vertical industries, proposes lifecycle standardization specifications for the collection, cleaning, annotation, and performance quality evaluation systems of representative datasets required for fine-tuning training.

9.2 Industry-Specific AI Applications

Defines dedicated AI system architectures and quantitative metrics for specific business workflows in sub-industries such as telecom services, Robotic Process Automation (RPA) development, power grid inspection diagnostics, logistics, and smart space home service robots.

ITU-T F.746.16: Technical requirements and evaluation methods of intelligence levels of intelligent customer service systems (2022)
For smart customer service systems involving multi-turn voice and text interactions, establishes an intelligence grading evaluation system and testing verification use cases based on AI core technology maturity and business intervention experience.

ITU-T F.748.71: Requirements and functional architecture of AI model generalization system for telecommunication intelligent customer service (2025)
Targeting telecom network O&M and fault processing businesses, standardizes the reference architecture of an AI customer service foundation model generalization system capable of utilizing fused datasets for fine-tuning to quickly adapt to communication scenario tasks.

This use case below describes the process of model generation of TICS for network operation and maintenance (O&M). The O&M-related customer services are mainly used to assist telecommunication employees to complete the strategy design, to maintain the stable operation of the network. The O&M-related customer services receive operation and maintenance requests (voice, text, image or video) from telecommunication employees and returns the corresponding recommended solutions.

F.748.71 Example of model generalization for intelligent network O&M customer service

ITU-T F.748.55: Technical requirements and evaluation methods of robotic process automation systems (2025)
Proposes detailed technical standards for new-generation Robotic Process Automation (RPA) systems integrating CV and NLP capabilities across dimensions like visual template development, component log monitoring, and application exception scheduling.

ITU-T F.747.11: Requirements for intelligent surface-defect detection service in industrial production lines (2022)
Formulates service model requirements and anti-misjudgment measurement metrics for applying machine vision and image segmentation technologies to perform smart surface flaw (e.g., scratches, pits) detection on industrial assembly lines.

ITU-T F.747.12: Requirements for artificial intelligence based machine vision system in smart logistics warehouse (2022)
Standardizes the capability requirements for machine vision system hardware and algorithm management components targeting smart logistics warehouse sorting, automated palletizing inventory, and abnormal barcode label recognition.

ITU-T F.748.65: Requirements and framework of AI-based cognitive inference system for multimedia applications (2025)
Establishes a cognitive reasoning and parsing system framework based on multimedia information, using spatiotemporal sequence deconstruction of relevant entity elements in audio/video data for causal logical Q&A responses.

ITU-T F.748.41: Technical requirements and evaluation methods of artificial intelligence (AI)-based driver behaviour detection application (2025)
Establishes workflow frameworks and measurement metrics for in-vehicle visual recognition AI applications (e.g., detecting driver fatigue states and alarm mechanisms) in IoV assisted driving scenarios.

ITU-T F.AIASSPDN-RF: Framework of AI-enabled analysis service system for power distribution network (2025)

Defines an intelligent pre-processing service and cause reasoning cloud platform architecture applying AI technologies to process engineering text drawings and electrical condition features in power distribution network systems.

ITU-T F.DCOR-Reqs: Requirements for construction of ontologies knowledge base for household service robots in intelligent spaces (2024)
Guides the standardized construction of knowledge base models required for home service robots' perceptual decision-making and hardware component data interaction in smart indoor spaces using "ontology" models.

ITU-T F.DCHU: Requirements for construction of home user profile ontology knowledge base for robot interaction in intelligent space (2025)
Standardizes the construction hierarchy and privacy authorization boundaries of home user profile ontology knowledge bases relied upon by service robots to provide personalized secure interaction services in smart home spaces.

ITU-T F.DCHE-reqs: Requirements for construction of ontology knowledge base of home environment for robot interaction in intelligent space (2025)
Addressing complex and variable environmental element states in home interiors, provides a home space environmental ontology graph knowledge base construction architecture guiding robots to adaptively establish interaction relationships with the environment and objects.

ITU-T F.MFMP: Requirements and framework for the meteorological foundation models platform based on multimedia data (2025)
Proposes software architecture requirements for application development, training, and intelligent forecasting deployment platforms of vertical meteorological foundation models integrating multi-modal data like radar and satellite cloud maps.

ITU-T F.REAIOCR: Requirements and evaluation methods for AI-based Optical Character Recognition Service (2020)
Standardizes the basic and enhanced functional assessment and security level requirement systems for AI Optical Character Recognition (OCR) services in card, bill, and natural scene text information extraction tasks.

ITU-T F.IMSS-AI: Framework and Requirements of Intelligent Marketing Service System based on AI (2025)
Defines the reference framework and scenario operational requirements for next-generation smart marketing customer service hub systems collaborating "general knowledge LLMs" and "professional visual assistance" capabilities.

ITU-T F.748.32: AI-based detection framework (2024)
Establishes framework requirements for intelligent word segmentation and review detection systems targeting illegal/sensitive text, images, and audio/video elements in short message services.

ITU-T F.748.33: Evaluation process of image-based Re-ID algorithm (2024)
Standardizes dataset scale calibration, video formats, and general evaluation workflow operational guidelines for evaluating the performance of image-based target Re-Identification (Re-ID) algorithms.