A consensus is investigated for two-dimensional nonlinear multi-agent systems with system faults, where the system input powers of each follower are unknown in this chapter. Based on adaptive control, fuzzy logic systems, and other control techniques, a novel distributed fault-tolerant control scheme with supervisory controllers is constructed such that the consensus is obtained in the presentence of system faults. Unlike the works in literature where system input powers equal one, the powers investigated in this chapter not only are larger than one but also are unknown. In sense of Lyapunov theory, it is proven that the synchronization among the agents are obtained and each synchronizing error with other closed-loop signals converges asymptotically to a small neighborhood of the origin. Finally, for the control scheme in this chapter, its effectiveness is demonstrated by simulation results.