The paper proposes a novel target following solution for an electric powered hospital bed. First, an improved real-time decoupling multivariable control strategy is introduced to stabilize the overall system during its operation. Environment laser-based data are then collected and pre-processed before engaging a neural network classifier for target detection. Finally, a high-level control algorithm is implemented to guarantee safety condition while the hospital bed tracks its target. The proposed solution is successfully validated through real-time experiments.