AI for Vehicle: advancing telecom and automotive collaboration through multi-agent systems for 5G B2B services

Commercial context

The 5G B2B services market is experiencing explosive growth, as 5G becomes a core engine driving industrial digital transformation. Service development however still faces significant challenges, such as slow service rollout, low operational efficiency, and issues with reliability. To overcome these bottlenecks, operators need new technical approaches to scale their 5G B2B offerings.

One key obstacle lies in manual service configuration. 5G service provisioning still heavily relies on manual operations, leading to inefficiency and risk. Handling complex network structures, large volumes of parameters, and frequent configuration activities remains a challenge. Inefficient alarm and complaint handling poses another major problem. Current systems for alarm and complaint handling are simply inadequate – traditional manual operations are too slow and resource-intensive. This inevitably and directly impacts SLA performance.

Solving these challenges presents a compelling commercial opportunity. Operators that can automate their 5G B2B workflows end-to-end will gain a significant competitive edge, offering faster delivery, reduced OPEX, improved reliability, and the ability to provide SLA-backed services. Moreover, this creates new revenue models and fosters long-term partnerships.

The solution

The Multi-Agent Boost for 5G Services Catalyst is key in enabling the seamless integration of AI technologies across multiple domains, helping telecom operators scale their 5G B2B offerings effectively. The project’s innovations span the entire 5G B2B service lifecycle, addressing both operational and strategic needs. By integrating initiatives like AI for Product, AI for Vehicle, AI for RAN, AI for VRC (Vehicle-Road-Cloud), and OTAI (Open Telecom AI Interoperability), the solution offers a holistic, AI-driven framework that supports connected vehicle ecosystems, autonomous networks, and the broader telecom industry.

The Catalyst’s key innovations include:

1. Quick and slow thinking models: The integration of quick thinking and slow thinking models allows for efficient processing of straightforward cases, while more complex scenarios can be handled with deeper inference. This resolves the traditional efficiency-precision paradox in telecom operations.
2. Knowledge graphs: The system uses knowledge graphs to map the dependencies between configuration instructions, parameter relationships, and auditing requirements. This ensures efficient and accurate service subscription processes and supports intent-based work order generation.
3. Chain-of-thought reasoning: Chain-of-thought reasoning is employed to automate alarm handling and complaint resolution. This allows for intelligent decision-making and problem-solving, significantly improving operational efficiency.
4. Signalling model: A new signalling model, based on transformer architecture, is used to automatically parse complex signalling messages, enabling automated network management and ensuring system integrity.

This Catalyst’s work spans a series of collaborative efforts. At the heart of the Catalyst is the AI for Vehicle initiative, which focuses on improving the connected vehicle ecosystem. By deploying AI-driven agents, it automates key functions like complaint handling and fault resolution, helping automotive manufacturers deliver more efficient, reliable services to their customers.

Meanwhile, the AI for Product initiative serves as the foundation for creating intelligent telecom products across various domains. It provides a scalable framework for AI integration, ensuring that AI solutions are built with a product-centric approach that can be deployed across different industries, including automotive and telecom.

The AI for RAN project optimizes Radio Access Networks (RAN) by integrating AI capabilities to automate network operations. AI models enhance network performance, manage resources efficiently, and enable real-time optimization, ensuring that 5G services are delivered with improved speed and reliability.

The AI for VRC initiative enables seamless communication between vehicles, road infrastructure, and cloud systems. This project focuses on creating an intelligent ecosystem where vehicles, roads, and the cloud are interconnected, allowing for better optimization of transportation networks and enhancing the overall connected vehicle experience.

And, finally, the OTAI project focuses on developing an open framework for telecom operators to enable AI-based automation and orchestration across their networks. By using OTAI standards, this project builds a robust, AI-native multi-agent architecture that supports the seamless integration of AI into telecom services, automating workflows and improving network performance.

The first official publications of AI for Vehicle, AI for RAN, and AI for VRC are scheduled for DTW 2026. These publications will mark significant milestones in the evolution of AI technologies within the telecom industry, providing critical insights and frameworks for the broader deployment of AI in real-world, connected environments.

Application and wider value

This Catalyst demonstrates how AI, autonomous networks, and multi-agent collaboration can drive telecom transformation. It proves that multi-agent systems can connect telecom and automotive ecosystems, enabling self-healing networks and monetizable services in the autonomous era.

Benefits for CSPs include lower OPEX, with predictive assurance and automation leading to 20–25% reduction in operational costs. There are also major gains to be made in terms of time-to-market, allowing new IoV services to be introduced 75% more quickly. The project team also forecasts a 60% reduction in repair times, alongside new revenue opportunities through monetization of AI-powered services such as network-as-a-service and V2X SLA products. Vehicle enterprises also benefit from proactive optimization, as rapid issue resolution and proactive system optimizations improve quality of service for connected vehicles.

More broadly, this Catalyst provides a blueprint for scaling AI and multi-agent collaboration across domains such as logistics, energy, and smart cities. It paves the way for AI monetization and sustainable mobility, benefiting society with reliable connectivity for over 23 million vehicles, promoting safer, greener mobility, and a more connected world.