China Mobile Releases Spatiotemporal Intelligence Series Model Capabilities and Benchmark Scenarios

Today, the 2026 Global Developer Pioneer Summit (GDPS) and the International Embodied Intelligence Skills Competition officially landed in Pudong. During the event, China Mobile held the Spatiotemporal Intelligence Empowering Advanced Manufacturing Development Forum under the theme "Intelligence Shaping Spatiotemporal, Gathering Momentum for a New Journey," and unveiled a series of spatiotemporal intelligence model capabilities and benchmark scenarios.

It is reported that this model capability is the result of China Mobile Shanghai Industrial Research Institute integrating its technological accumulation in fields such as 5G+AI+BeiDou. Closely aligned with the practical needs of the digital and intelligent upgrading of manufacturing, it has built a core spatiotemporal intelligence capability system covering three major technical routes—"spatiotemporal reconstruction, spatiotemporal understanding, and spatiotemporal prediction"—and innovatively constructed an L1-L4 four-tier scenario architecture, with capabilities covering multiple key fields of the real economy, including natural resources, industrial production, urban transportation, and intelligent transport. At the L1 terrain level, the model focuses on large-scale geophysical remote sensing scenarios and can achieve refined extraction of surface feature elements, dynamic detection of multi-temporal surface changes, and intelligent semantic classification of aerial survey point clouds, providing reliable technical support for scenarios such as natural resource exploration and precise monitoring of surface ecological environments. At the L2 city level, the model creates a virtual-real integrated dynamic real-scene digital twin, completes spatiotemporal correlation perception and simulation preview of operations for park equipment and environments, and realizes digital control of all elements and the entire process in parks. At the L3 regional level, relying on multi-channel video real-time 4D scene reconstruction and multi-target continuous spatiotemporal tracking capabilities, the model accurately assesses intersection traffic flow, traffic conflict risks, and congestion evolution trends, providing intelligent decision support for industrial logistics transportation and factory traffic scheduling. At the L4 local level, the model focuses on intelligent cockpit scenarios, connecting spatiotemporal fusion perception inside and outside the cabin with contextual intelligent decision-making, upgrading traditional passive human-vehicle interaction to proactive in-vehicle services, and comprehensively empowering experience upgrades for intelligent manufacturing equipment and intelligent transport units. During the forum, China Mobile, together with multiple industry chain partners such as Inspur Group, Guangxiang Technology, Chicheng Technology, and Datatang, held a spatiotemporal intelligence ecosystem signing ceremony, joining forces to build a spatiotemporal intelligence ecosystem featuring "joint business expansion, joint partner building, and shared industry prosperity."

Spatiotemporal intelligence is happening now. China Mobile Shanghai Industrial Research Institute will continue to iterate and improve its technological foundation, join hands with all sectors of the industry to continuously expand the application boundaries of spatiotemporal intelligence, and provide solid technical support for the high-quality development of the digital economy and the construction of a smart society.