AI-Based Performance Analysis of Temperature Sensors A Reverse Engineering Case Study

Abstract

To replace imported Resistance Temperature Detectors (RTDs) in heavy duty engines such as those found in power plants construction and mining machinery and tracked vehicles this thesis describes the development and validation of a locally made temperature sensor. The study aims to lessen reliance on expensive imported sensors that are susceptible to interruptions in the supply chain. The study created a prototype utilizing locally sourced materials by reverse-engineering a typical automotive coolant sensor. To validate the sensors performance in real time a complex test bench was built and outfitted with cutting edge data acquisition systems and artificial intelligence algorithms. In terms of response time and stability under different conditions, in particular the results showed that the native sensor matched the accuracy and dependability of imported RTDs. Performance analysis was improved by using AI in the validation process giving rise to the possibility of using AI in sensor development in the future. As a first step toward incorporating artificial intelligence (AI) into the verification of critical components this thesis provides a workable substitute for a crucial import dependency and opens the door for further advancements in sensor technology.