Dismantling the three-layer capability structure of Mofa Technology’s AI screen

With the rapid development of embodied intelligence technology, display terminals are gradually evolving from traditional information display carriers to human-computer interaction portals with perception and interaction capabilities. "Screen intelligence" has become an important direction of industry concern. Under this trend, the integration of AI and display technology continues to deepen, promoting the upgrade of screens from "display tools" to "smart terminals".

As one of the early start-ups in the field of embodied intelligent digital humans, Mofa Technology is also a leader in financing in this field. More importantly, the company has realized the deployment and coverage of digital human capabilities on multiple types of terminal screens, forming a unified multi-terminal capability covering large screens and small and medium-sized screens.

At the 2026 TrendForce New Display Industry Symposium (DTS 2026), Professor Chai Jinxiang, CEO of Mofa Technology, delivered a keynote speech "Upgrading Every Screen to an AI Screen". He systematically introduced the application progress and implementation practice of AI digital humans on multi-terminal screens, and elaborated on the development path for the evolution of screens into embodied intelligence.

Professor Chai Jinxiang, CEO of Mofa Technology

AI-driven display upgrade, the screen enters the smart service entrance stage

In the context of the current continuous evolution of digitalization and intelligence, display terminals have widely covered various scenarios such as LED large screens, LCD screens, self-service terminals, mobile phones, computers, vehicle screens, and AR/VR equipment. With the development of AI technology, the functional boundaries of screens are being redefined, and its core trend is shifting from "display tools" to "AI screens".

The essence of the so-called AI screen is to upgrade the screen into an intelligent system with interaction and service capabilities through large models and intelligent agent capabilities. Under this system, whether it is a large screen in the exhibition hall or a car central control screen, it will be upgraded from information display equipment to an intelligent terminal with service capabilities. In the future, as the penetration rate of smart terminals increases, this trend will further accelerate.

In this evolutionary path, the role of the screen is fundamentally changing. AI screens can be understood as embedding intelligent agents with complete capabilities into various display terminals, transforming them from "device" to "service subject". Regardless of LED screens, LCD screens, TVs or mobile terminals, they can all host digital employees with interactive capabilities to achieve online and standardization of service capabilities.

From the technical nature, embodied intelligence not only has a "brain" but also a "body". The large model provides cognitive and decision-making capabilities, while the screen becomes the carrier of execution and interaction, which is also equivalent to the "body" of the intelligent agent.

Embodied intelligence empowerment, screen capabilities evolve into service-oriented agents

When the screen has embodied intelligence capabilities, its capability structure can be compared to a complete service-oriented individual: on the one hand, it has cognitive and task capabilities based on large models; on the other hand, it has the ability to complete business execution through peripherals and systems; it also has visual and speech perception capabilities; and it can realize multi-modal expression of voice, movement and expression through digital human images.

Through the combination of cameras, microphones and AI perception systems, the screen can achieve real-time understanding of users and the environment; through access to large models and intelligent agent systems, task execution and service scheduling can be achieved; through digital human expression capabilities, human-like interactive experiences can be achieved. This transforms the screen from a passive display device to an active service terminal.

From the perspective of application value, AI screens can directly assume the role of service entrance based on the capabilities of embodied intelligence. For example, it serves as an intelligent explanation and inquiry assistant in the exhibition hall scene, as an AI housekeeper in the hotel scene, as a shopping guide assistant in the business scene, and provides reception services in the front desk scene, realizing the transformation from "display media" to "service employees".

This change will also reshape the industry structure. The introduction of AI screens will superimpose software and service capabilities, expanding the industry from a single hardware market to a composite market of "hardware + AI + services". At the same time, application scenarios will also expand from relatively fixed display spaces to wider fields such as medical care, government affairs, transportation hubs, and commercial retail.

Three-layer capability architecture to build the core technology system of AI screens

From the perspective of technical implementation path, AI screens are mainly composed of three levels: the first is the perception layer, which realizes environment and user recognition through visual and voice capabilities; the second is the cognitive layer, which realizes understanding, reasoning and task execution through large models and agent systems; the third is the expression layer, which realizes the natural output of voice, movements and expressions through digital human technology.

At the level of expression, core capabilities mainly include AI human-creation capabilities and AI embodied driving technology. Among them, AI human-creation technology can build more than 3,000 hyper-realistic 3D high-quality character libraries, support personalized editing, and realize the generation of 3D digital humans from images with diversified styles, effectively reducing the use threshold and production costs, and improving large-scale production capabilities. In terms of AI embodied driving, Mofa Technology has built the world's first Vincent 3D multi-modal large model, as well as the world's first AI-end rendering and end-side solution system.

Relying on the above-mentioned core technology advantages, Mofa Technology's embodied intelligent digital human has six major capabilities: high quality, low latency, high concurrency, low cost, multi-terminal adaptation and personalization. At the same time, the company has created Mofa Nebula - an open platform for embodied intelligent digital people (PC and mobile), and the "Mofa Youyan" platform, gradually building a complete product and capability system.

In addition, the perception layer capabilities are based on multi-modal perception and realize functions such as identity recognition, speech recognition, visual recognition, expression recognition, and behavior and environment recognition. This part of the capability will also be simultaneously launched on the company's platform.

Cognitive layer capabilities are reflected in the ability to "have a brain and be able to do things": after users upload files, the large model can process knowledge and complete knowledge management, while supporting the editing of agents and one-click deployment to the terminal. At present, the "Mofa Youyan" product has supported knowledge uploading, knowledge understanding and knowledge application, forming a complete closed loop of capabilities.

Multi-scenario implementation verification, accelerating the process of screen intelligent application

Judging from the actual implementation, Mofa Technology-related applications have been launched in multiple scenarios. There are currently several cooperation cases, including the implementation of applications in hospital terminal scenarios with Vislu; a strategic cooperation with Unilumin Technology in the field of LED display to promote the upgrade of traditional LED screens to embodied intelligence; in LCD, LED screens and AGV mobile device scenarios, it cooperates with Reget to develop integrated terminals to achieve navigation and interactive functions, and to support real-time Q&A and service capabilities during the mobile process.

In addition, Mofa Technology has also cooperated with TPV to launch digital micromarket products, which realize product shopping guide and inquiry service functions through cartoon digital images; it has carried out relevant application exploration with TCL in hotel TV scenarios; related capabilities have also been extended to mobile terminal devices, covering various application scenarios such as medical care, retail, transportation guides and hotels, which initially reflects the implementation of screen intelligence.

From the perspective of industrial structure, the traditional screen industry focuses on display functions, but driven by AI, screens will gradually have a capability structure similar to that of an operating system, forming an embodied intelligent system and application ecosystem. In this system, intelligent agents such as digital humans, shopping guides, front desks, and government services will gradually be standardized to build a new application ecosystem.

Overall, screens are evolving from "display devices" to "AI service terminals". When the screen is endowed with the ability of perception, cognition and expression, its essence will no longer be an information carrier, but an intelligent agent that can independently provide services.

Looking to the future, Mofa Technology hopes to make every screen an embodied intelligence, enabling it to have real interactive capabilities and service capabilities, thus reconstructing the connection between people and information, and between people and services. (Text: LEDinside Mia)

Please indicate the source when reprinting! For more LED information, please pay attention to the official website (www.ledinside.cn) or search the WeChat public account (LEDinside).