CNCC | Prof. Fei Wang successfully organized the themed forum about New Paradigms for Complex System Research Driven by Spatiotemporal AI at CNCC 2026!

CNCC | Prof. Fei Wang successfully organized the themed forum about “New Paradigms for Complex System Research Driven by Spatiotemporal AI” at CNCC 2026!

Spatiotemporal complex systems are widely found in natural, social and engineering domains. Their evolution underlies critical scientific and engineering challenges including Earth environment, urban operation, ocean dynamics, energy security and public health, making them a key research target for AI4Science targeting scientific discovery and complex decision-making. In recent years, centered on spatiotemporal foundation models, multi-task pre-training, benchmark evaluation, reinforcement learning and other directions, spatiotemporal artificial intelligence has been gradually building shared foundational capabilities across diverse fields. This shift advances complex-system research beyond isolated prediction tasks toward unified modeling, complex reasoning and autonomous decision-making. Nevertheless, when applied to real-world scientific problems, spatiotemporal AI still faces major bottlenecks: insufficient adaptability of fundamental network architectures, limited cross-task generalization of pre-trained models, difficulties in establishing unified and fair benchmark suites, and inadequate integration of physical laws, domain knowledge and AI models. Centered on the theme From Prediction to Discovery: New Paradigms for Complex System Research Driven by Spatiotemporal AI, this forum will deliberate on the above topics. Focusing on representative scenarios such as meteorological forecasting, ocean eddy analysis, smart cities and complex network optimization, it will explore the deep fusion of general AI approaches and domain knowledge, and advance complex-system research from prediction toward cognition, decision-making and scientific discovery.