Identify Abnormalities

Detect vessels displaying abnormal behaviors at sea to support in revealing unauthorized or illegal maritime activities.

Which unusual vessel behaviors can you detect with OceanIntel?

Below, you’ll find the abnormal vessel movements that OceanIntel’s cutting-edge algorithms can identify:

  • Activation and deactivation of AIS transponders
  • Deviation from expected route
  • Deviation from normal sailing patterns (speed/course changes etc.)
  • Detection of spoofed positions
  • Identification of messages that are supposed to be present 
  • Coming soon: identification of rendezvous events (ships approaching each other)

Discover how you can leverage this advanced solution to gain insights within your maritime domain which are typically hard to obtain through other means.

What is the technology behind identifying abnormal vessel movements?

Utilizing extensive historical AIS data collected over nearly two decades within the EU and a decade worldwide, we’ve successfully established and trained algorithms designed to identify unusual behaviors. The following outlines the functions performed by these algorithms:

1. Based on all historical data, find normal behavior:

  • Find typical route for ship X from A to B.
  • Where does ship X normally sail to / which area is it normally in / how does it normally move (speed/course changes etc.)?
  • What will ship of type Y, size Q normally do to get from A to B?
  • For ship X, get prediction for next port.

2. Based on the above, look for anomalous behavior. Any behavioral pattern that is inconsistent with the recorded typical behavior is flagged as anomalous behavior. 

3. If anomalous behavior occurs, predict ship position in near future:

  • 1-2-3 hours out, where is ship expected to be?
  • Take a satellite image (SAR/Image)


Do you want to explore additional behaviors that haven’t been discussed yet? Let’s engage in a conversation to uncover untapped possibilities. We are always eager to develop the most cutting-edge solutions imaginable.