WAVE Powers a Live Service Demonstration at a Leading Global Forum
This year’s Summer Davos focused on “Innovating at Scale,” with artificial intelligence and the application of emerging technologies in the real economy among the forum’s key topics. A key question for the industry is how embodied intelligence can move beyond laboratory demonstrations and achieve commercial deployment at scale.
Moving beyond conventional static hardware displays, Matrix Robotics created a live demonstration area in the venue’s first-floor dining zone. The service solution is built around the WAVE Embodied AI Foundation Model and a general-purpose full-body motion control system. Combining cross-scenario generalization, stable execution of long-horizon tasks and high-degree-of-freedom dexterous hands, the system provided live coffee service to international attendees and demonstrated its ability to operate in a real-world service environment.
The WAVE model handles perception, reasoning and high-level motion planning, while the full-body motion control system coordinates the robot’s physical movements. Long-term testing across multiple scenarios has demonstrated stable performance, enabling the system to support frequent and continuous service at the venue. MATRIX-3 serves as the physical platform through which the intelligent system performs real-world tasks.
Generalization Enables Autonomous Coffee Service, While Dexterous Hands Demonstrate Precise Manipulation
In the dedicated demonstration area, the WAVE Embodied AI Foundation Model coordinates the workflow together with the full-body motion control system and high-degree-of-freedom dexterous hands, enabling MATRIX-3 to perform a standardized coffee-service process. Using the model’s multimodal visual perception, the robot independently identifies and picks up an empty coffee cup, calculates the force required to press the coffee machine’s controls, lifts the hot beverage after brewing, plans a route to the guest and delivers the finished drink steadily.
The WAVE model incorporates physical force-control reasoning, allowing the system to adjust the dexterous hands’ grip strength in real time and reduce the risk of cups slipping or hot beverages spilling. The continuous, multi-step workflow—from picking up a cup to delivering a hot beverage—demonstrated the WAVE model’s multitask processing and generalization across different movements, highlighting the role of the underlying AI system in complex physical interactions.
WAVE Supports Replicable Solutions for Commercial Service Applications
As the forum’s sole humanoid robotics service provider, Matrix Robotics used live coffee service to demonstrate its approach to a key industry challenge: many humanoid robots emphasize hardware demonstrations, while limitations in model generalization make rapid deployment across multiple scenarios difficult.
Matrix Robotics has focused on deploying robotic systems in standardized, repetitive service environments, with its core technology centered on the WAVE Embodied AI Foundation Model and general-purpose full-body motion control system. With multitask adaptability and cross-scenario transfer capabilities, the system demonstrates its potential for deployment at international forums, exhibition venues, retail stores, cultural and tourism spaces and other service environments, offering a replicable solution for the digital transformation of the global service industry.
International Attendees Experience China-Developed AI Technology Firsthand
Throughout the forum, entrepreneurs, technology professionals and Chinese and international media visited the dining-area demonstration to experience the autonomous coffee service powered by the WAVE Embodied AI Foundation Model and performed by MATRIX-3’s dexterous hands. Attendees received coffee prepared by the robot and observed the system’s stable and fluid operation firsthand.
A senior executive of Matrix Robotics said that Summer Davos provides a leading global platform for showcasing China’s AI innovation. The live coffee-service demonstration provides a direct example of the company’s commercialization strategy centered on the WAVE model. Matrix Robotics independently develops the WAVE Embodied AI Foundation Model, the general-purpose full-body motion control system and high-degree-of-freedom dexterous hands. The company continues to improve multitask generalization, safe human-robot interaction and extended continuous operation, with the goal of moving Physical AI from concept demonstrations toward commercial deployment at scale.
Expanding Applications and Accelerating Embodied AI Deployment at Scale
As the global embodied AI industry enters a period of rapid development, general-purpose AI model capabilities and commercialization across different scenarios are becoming key areas of competition. The forum’s theme, “Innovating at Scale,” closely aligns with Matrix Robotics’ long-term development strategy centered on the WAVE model.
Combining the WAVE model’s general-purpose and long-horizon task capabilities with high-degree-of-freedom dexterous hands, the integrated system can support applications beyond coffee service, including exhibition reception, material handling and precision object manipulation, creating broad potential for deployment across industries.
By participating in Summer Davos in Dalian, Matrix Robotics demonstrated the real-world operating capabilities of its independently developed intelligent system and created a platform for international industry exchange and cooperation. Going forward, Matrix Robotics will continue to advance the WAVE Embodied AI Foundation Model, the general-purpose full-body motion control system and general Physical AI algorithms, while expanding applications across retail, cultural tourism, exhibitions and industrial support. With foundation-model technology at the core of its strategy, the company will work with global industry partners to accelerate the large-scale deployment of embodied AI and bring China-developed Physical AI technology to a broader global audience.
Hashtag: #MatrixRobotics
The issuer is solely responsible for the content of this announcement.
