Maritime Adaptive Automation

TNO developed an Intelligent Operator Support System for operators working with Dynamic Positioning Systems. The incentive is the trend of an ever increasing level of automation and, with that, a decline of manual control in case something goes wrong. Humans are not particularly well suited for a supervisory role and in maintaining a high level of Situational Awareness needed for dealing with complex control issues in highly automated domains.
The DP system keeps the vessel above the oil well, despite wind and currents

Most offshore operations, like dredging, trenching, pipe and cable laying, use Dynamic Positioning (DP) systems. DP Systems automatically hold a vessel’s position or sailing course. The main function of an Floating Production and Offloading Facility is to pump oil. The DP system keeps the vessel right above the oil well, despite wind, waves and currents.

The DP system is in control most of the time. However, four operators working in shifts must supervise the system 24/7 to prevent possible drift-off. The operator in charge must remain focused. Research shows that can be a challenge, especially at night. Studies also show, that operators often don’t have enough situational awareness to make important decisions quickly and effectively. Also, since proficient operators are hard to find, it’s important to make their jobs more interesting and to do more, with fewer people.

IOSS built on top of the existing DP control loop

That’s why TNO developed an Intelligent Operator Support System (IOSS). Built on top of the existing DP control loop and alarm handling functionality, IOSS:

  • Supports for the early detection of critical conditions
  • Provides on-time and context specific information for just in time situational awareness
  • Provides procedural support based on checklists and Activity Specific Operating Guideline’s.
  • Allow operators to leave the bridge and perform other tasks, when conditions allow.

For detection of critical conditions, IOSS uses A.I. techniques like predictive analytics to recognize trends in the data, and uses semantic technology to communicate findings to operators in a meaningful dialogue. The recognizable IOSS interface engages in conversations with the operator, and functions like a virtual teammate.

Explainable A.I. is what sets IOSS apart from traditional alarm based systems. Unlike an alarm, which contains a brief text statement about a problem, IOSS interacts with operators with smart notifications, which presents the message in layers that can be exposed using a dialogue. Explainable A.I. enhances the transparency, understandability and predictability of complex automated systems. This type of interaction becomes more and more relevant as computers are applied to more difficult problems for which an adequate solution cannot be established by one actor alone.

IOSS also contains human aware computing, because not only humans must be able to understand automation, also automation must be able to model the state of the operator and take that into account when providing support. A simple example is using an eye tracking device to determine whether or not the operators sits behind the workstation. location tracking can be used to determine the distance between the operator and the bridge and take this into account when warning the operator.

Currently the IOSS is a prototype to demonstrate that IOSS can meet the needs of offshore operations using DP operator in the loop scenario’s. TNO would like to bring IOSS to a higher technology readiness level. For this, we are seeking collaboration with industries.

Downloads

Early Research Program: Human Enhancement

Dr Hans van den Broek

Contact

Dr Hans van den Broek

Email

We use anonymous cookies to enhance the use of our site.