The paper examines the development of a digital twin for diagnosing the condition of mechanical cleaning devices for pump-compressor tubes. The issue arises from the formation of asphalt-resin-paraffin deposits (ARPD) on the walls of pump-compressor tubes and other equipment during oil well operation. The intensity of this process sharply increases with decreasing reservoir pressure and oil temperature. To remove ARPD, wells are periodically cleaned, most often mechanically using specialized scrapers. To improve the efficiency of mechanical cleaning devices, the authors propose implementing a digital twin for diagnosing the condition of key components. This system analyzes real-time parameters of the device, monitors scraper condition, and predicts potential emergency situations. The digital twin employs a mathematical model based on Mamdani’s fuzzy logic algorithm. Testing on an operational well at the Buzovyazovskoye field confirmed the model’s adequacy, demonstrating diagnostic accuracy of approximately 80%.