The present paper tackles the problem of the stability in autonomous bicycles. A conventional PI controller has been designed for different values of the system velocity. An adaptive controller has been designed to optimize the PI gains at different system velocities to minimize the energy consumption of the control system for disturbance rejection. The simulation results show that the proposed adaptive control strategy provides notorious energy consumption saving and guarantees the lateral equilibrium of the control system in presence of external disturbances. Finally, a case of study that applies the proposed control to trajectory tracking using a dynamics inversion technique is detailed.