Induction machines, crucial in various industrial sectors and the economy, are sought after for their dependable and secure functioning. Their faults or malfunctions can cause prolonged downtimes and significant financial losses, driving the need for continuous online condition monitoring. This monitoring involves real-time measurements during operation to swiftly identify faults, thereby mitigating unexpected failures and reducing maintenance expenses. This paper commences by scrutinizing the typical operation of asynchronous motors, followed by an examination of their behaviour under abnormal conditions such as stator or rotor winding defects. Through theoretical analysis of breakdown scenarios, parameters linked to emerging defects are identified, guiding the selection of appropriate detection methods during motor operation. The alignment between theoretical analysis and experimental data from the Laboratory of Electric Machines reinforces the findings of this study.