Commonly used industrial controllers are designed to operate on one of the control modes-Proportional (P) mode, Proportional Integral (PI) mode, Proportional derivative (PD) mode, and Proportional-Integral-Derivative (PID) mode. However, every control mode has its own merits and demerits. When the system parameters are altered, the performance of the selected controller may deteriorate rapidly. To overcome this issue, intelligent online tuning of controller settings is proposed by several researchers. Although this method may improve the controller performance to some extent, it may be far below than that of best fit control mode for that condition. Hence, this work provides a methodology to evaluate the performance of different control modes and switch it to the most appropriate control mode in a real time manner automatically. Effectiveness of the proposed approach is demonstrated through an example of self-balancing robot.