This paper develops an autonomous-driving platform to check the condition of soil in agricultural areas. The platform, which runs on an electric motor equipped with a caterpillar, uses an RTK-GPS module for accurate position estimation to obtain position coordinates with an accuracy of several centimeters and integrates an Inertial Measurement Unit (IMU) sensor to obtain precise angle values. An algorithm is developed to control the autonomous-driving platform and proceed with the planned route to the optimal route. In order to smoothly and accurately control the direction of a small tractor, an extended Kalman filter control algorithm is used to develop an algorithm that can automatically achieve more stable path driving by eliminating instability and risk caused by rapid steering control. The route is set by combining the RTK-GPS module and IMU module, which provide precise location coordinates, and an environment is provided so that the soil contamination sensor can operate at an accurate location at regular intervals within the driving route.