Unstable approaches are often associated with hazardous and/or inefficient operational events at airports. They represent one of the most common contributing factors to runway excursions and anomalous flight trajectory behaviour. The pilot, upon identifying that the aircraft is not in a position for landing, can perform a go-around and/or a missed approach procedure. In dense terminal control areas, this procedure can generate additional workload for the air traffic controllers in this reduced volume of airspace, as it requires a quick and accurate reaction in order to reintegrate the flight into the traffic flow and immediately establish a new approach and landing sequence. In current operations, air traffic controllers typically become aware of a go-around or a missed approach during or after their execution, which leads to a very short time horizon for decision-making. This work aims to propose a novel real-time monitoring and alerting solution to increase the situational awareness of air traffic controllers in approach control facilities and aerodrome control towers. The solution is based on the use of surveillance data and machine learning algorithms for identifying flight path anomalies in real-time and predicting the execution of a go-around. The algorithms would generate an alert to be presented at the Human Machine Interface in the air traffic control facilities, showing the exact moment when the aircraft enters an unstable approach behaviour, and the pilot could initiate a go-around.