This paper investigates the remote localization problem for a network-based automatic guided vehicle (AGV) system. Considering the constraints of network bandwidth and the interference of unknown-but-bounded (UBB) noises, a novel quantized set-membership approach is proposed. First, a differential zooming strategy is designed for the quantization process of the measurement signals in order to reduce the quantization error and the negative influence of the quantization effects. Then, a remote modified set-membership filter is designed to eliminate the influence of UBB noises and obtain the optimal position state estimation ellipsoid for the considered AGV. Sufficient conditions for the existence of the set-membership filter are derived and a recursive algorithm is provided for computing the ellipsoid that guarantees to contain the true state. Finally, the remote localization performance analysis verifies the effectiveness of the proposed quantized set-membership approach for the considered AGV.