Using the spontaneously synchronized clapping in a concert hall as a special case, we build a nonlinear emergent model to characterize the collective behaviors of the complex multi-agent systems. Based on this model, we develop an experimental platform for emergent computation, which only depends on the local interaction and reveals the uncertainty and diversity of emergent behaviors. By analyzing the data in the procedure of many hands clapping, we find that there exists an obvious critical region generated in the procedure from disorder to synchronization. More- over, we propose a fundamental synchronous criterion as follows： If the coupling coefficients c1 and c2 satisfy the condition 0.02≤c2=≤0.965c1＋0.018, then the clapping can realize synchronization.
