Abstract: Nonvisual sensors are essential and important for mobile robots. To recognize and understand the environment out of the robots themselves. The basic nonvisual sensors for wheeled mobile robots include sonar sensor, infrared sensor and tactile sensor. All these sensors are used in groups and their signal processing will occupy much CPU time. In this paper, we propose a multi-DSP-based system to control the sensors and process the corresponding feedback signals, thus the robot controller can be mainly used for its deliberative purpose, such as path planning and navigation. This paper presents the operation theory of signal processing for the above sensors respectively. Parallel processing methods are employed to accelerate the processing speed of the system. Experimental results show the effectiveness of the proposed method.