A vertical plane bi-dimensional strap-down inertial navigator with a redundant architecture is presented. Its detection unit is organized under the form of three inertial sensors arrays in linear redundant configurations. The inertial sensors data in each of the three arrays (two for accelerometers and another one for gyros) are fused by using a statistical method based on a Kalman filter. Shown are: the navigator inertial measurement unit structure, the fusion algorithm theory, the inertial navigator theory of operation, and the experimental validation of the proposed architecture. In the experimental validation phase the positioning, speed and attitude errors are evaluated and discussed relative to the reference signals provided by an integrated INS/GPS navigator.