The structural deformation generated by multi-source heterogeneous decommissioned 3C products in the service stage may cause the failure of automatic disassembly equipment. The performance drift failure will also disturb the disassembly process planning and production scheduling. Taking the decommissioned 3C product mixed-flow disassembly line as the research object， this study investigates the insertion order scheduling problem caused by decommissioned product disassembly failure in decommissioned products. An AND/OR node network model is constructed to quantify uncertain disassembly sequence and depth. With the optimization goal of maximizing workstation utilization and the variable of disassembly fault product buffering and transfer threshold， a mixed manual-automatic line reordering schedule model is established. Then， it is encoded by two-dimensional real-number coding and solved by a genetic algorithm. Finally， data instances of 58 varieties of decommissioned smartphones with mixed manual-automatic disassembly lines are selected for verification. The results show that the established manual-automatic hybrid line scheduling model can effectively develop the threshold scheduling scheme of maximum workstation utilization. The optimized maximum utilization rates of the three types of mixed-flow disassembly lines in the example clustering are 80.3%， 73.2%， and 81.3%， respectively.