Addressing major demands in communication, sensing, computing, and energy, we are dedicated to building next-generation intelligent network infrastructure that integrates these capabilities. Using the SRCON real-world network simulation engine as the technical foundation, we develop network modeling and intelligent optimization algorithms that combine data and mechanisms, advancing network operations toward autonomy. We also develop distributed intelligent sensing technologies to support low-altitude economy development and smart urban governance, construct cloud-edge-end collaborative computing network architectures for efficient cross-domain resource scheduling and optimization, and create specialized solvers for the energy sector to enable intelligent regulation across planning, operation, and recovery. Guided by the core philosophy of "data-driven, algorithm-empowered, and intelligence-native," we aim to build a comprehensive intelligent technology stack for network scenarios, serving as the core engine for infrastructure intelligence in the digital era.

Focusing on the deep integration of intelligent systems with physical environments, we aim to build next-generation machine intelligence entities with closed-loop "perception-decision-control" capabilities. Through hardware-software co-design as the core approach, we develop next-generation machine learning and deep reasoning algorithms, with a focus on enabling efficient deployment and adaptive optimization of intelligent algorithms on specialized chips, heterogeneous computing platforms, and edge devices. We develop lightweight, low-latency intelligent engines for edge and end-side applications, ensuring optimal alignment between algorithm performance and hardware computing power. At the same time, we explore cutting-edge directions such as embodied intelligence, multimodal perception, and physical interaction, driving machine intelligence from the digital realm into the physical world. Adhering to the core philosophy of "algorithms driving hardware, hardware enabling intelligence," we build full-stack capabilities from theoretical models to real-world systems, providing highly reliable, deployable, and self-evolving intelligent cores and physical implementations for autonomous unmanned systems.

Addressing the intelligent upgrade needs of industrial and consumer application scenarios, we are committed to accelerating the transformation and implementation of cutting-edge intelligent technologies into real-world products and systems. Guided by the principle of "full-stack collaboration and hardware-software integration," we leverage collaborative R&D teams spanning algorithms, software, hardware, and products to establish end-to-end engineering capabilities from data to products and from models to physical entities. We focus on industrial data governance and knowledge construction, promoting the development, deployment, and iterative optimization of intelligent agents in real-world scenarios. We also develop intelligent hardware R&D systems for consumer applications, enabling the design and mass production of embedded, low-power, and highly reliable edge-side intelligent systems. By establishing agile and platform-based engineering development processes, we create rapid iteration cycles from technical prototypes to mature products, striving to become the engineering hub connecting fundamental research and industrial applications. Our goal is to provide deployable, scalable, and sustainable intelligent solutions.