SHIPMA is not suitable for:

• detailed design;
• complicated mooring manoeuvres;
• training purposes;
• ship/ship interactions and;
• complex human interference.

For these purposes a study with a real-time manoeuvring simulator, possibly in combination with a SHIPMA study, is required. One important reason is that, because of the predictable character of the autopilot, SHIPMA runs can be reproduced exactly. This is of course in contrast with human control and the variability of wind, waves and current. Thus, it would not make sense to analyse SHIPMA results statistically.