In order to study the mapping and positioning effect of robot in indoor fishpond environment, a robot mapping and positioning system based on robot operating system (ROS) and sensor is designed. Firstly, the simultaneous localization and mapping (SLAM) system model of fishpond robot is established. Then, the sensor is used to extract the environmental features, and the mapping function based on Gmapping algorithm is realized. Under the condition of using the Adaptive Monte Carlo (AMCL) algorithm and selecting the basic experimental group of the fishpond with square corners, the positioning performance of the fishpond cleaning robot is studied and analyzed through the comparative experiments of different factors. The results show that the average positioning distance error is 9.02 cm and the average positioning angle error is 4.6 ︒ when the robot moves at a speed of 0.1 m/s. The research results show that the fishpond cleaning robot can construct high-precision map in unknown environment and achieve effective positioning, with good positioning and navigation ability.