This dataset is developed to support research on early warning of pilots' cognitive collapse. It encompasses the recordings of 10 participants over 3 separate sessions in the simulated flight experiment paradigm. The experiment aims to elicit the pilots' cognitive collapse state and to early warn of the tipping points, for details you can refer to the paper "An Early Warning Approach for Pilots’ Cognitive Tipping Points Based Multi-modal Signals". Each trial consists of 3 stages that can induce cognitive states at low, medium, and high level with a cognitive collapse.