This work provides an asymptotically stable and robust tracking control scheme using a finite-time stable disturbance observer in the feedback loop, for an unmanned vehicle modeled as a rigid body. The dynamics of the system is discretized using a Lie group variational integrator in the form of a “gray box” dynamics model that also accounts for unknown additive disturbance force and torque. These disturbance terms are estimated using the finite-time stable disturbance observer in real-time and then compensated by the control scheme. The stability analysis for translational and rotational motions is carried out separately. It is shown that the discrete-time control laws achieve asymptotically stable tracking of the reference position and attitude trajectories.