Page 18 - IJOCTA-15-2

Artificial intelligence-assisted station keeping for improved drillship operations
intelligence in three key aspects: First, it success- significantly reduces time delays compared to tra-
fully processes and correlates the complex rela- ditional GPS-dependent systems, eliminating the
tionships between structural forces and vessel re- communication lag associated with position refer-
sponse without explicit mathematical modeling. encing.
Second, it showcases remarkable predictive capa- By operating independently of GPS and
bilities by anticipating and counteracting distur- internet-based systems, this AI-driven solution
bance forces before they cause significant posi- enhances the vessel’s operational reliability and
tion deviation. Third, and most importantly, the robustness under extreme sea conditions. Inte-
algorithm achieves this while avoiding the com- grating AI with local control mechanisms further
mon pitfalls of over-correction or oscillatory be- streamlines the station-keeping process, demon-
havior that plagued previous control attempts. strating reduced deviations in position and ad-
The blue line’s stable trajectory, maintaining po- equate compensation for environmental forces.
sition within approximately ±3 m of the setpoint These advancements collectively improve the op-
(compared to the potential 160 m drift), vali- erational capabilities of the drillship, ensuring
dates the algorithm’s sophisticated understanding higher efficiency, reliability, and reduced down-
of the system dynamics. It represents a signifi- time in challenging offshore environments.
cant leap forward in marine control technology.
AI effectively bridges the gap between theoreti- Acknowledgments
cal control models and real-world environmental
complexities, offering a more robust and practical None.
solution for dynamic positioning systems.
Funding

None.

Conflict of interest

The authors declare no conflict of interest.

Author contributions

Conceptualization: Srinivasan Chandrasekaran
Formal analysis: Mahalakshmi Perala
Figure 23. Controlled response due to artificial in- Methodology: Mahalakshmi Perala
telligence Writing – original draft: Mahalakshmi Perala
Abbreviation: LCG: Longitudinal center of gravity
Writing – review & editing: Srinivasan
Chandrasekaran, Ermina Begowic

5. Conclusion Availability of data

The trained AI control algorithm calculates the Data is available upon reasonable request to the
force required to maintain the vessel’s position corresponding author.
and ensure effective station-keeping. By integrat-
ing the onboard INS, the position changes are de- References
tected accurately and without reliance on GPS or
acoustic-based systems. This approach enhances 1. Machado L do V, Fernandes AC. Moonpool di-
economic viability by eliminating the need for mensions and position optimization with Genetic
external positioning systems and providing real- Algorithm of a drillship in random seas. Ocean
time local control, ensuring prompt and efficient Eng. 2022;247.
http://dx.doi.org/10.1016/J.OCEANENG.2022.
corrective actions. The AI control algorithm dy-
110561
namically identifies the response data from the
2. Chandrasekaran S, Phoemsapthawee S, Krishna
INS system, rapidly retrieving the required mag-
S, Hari S. Fundamentals of Offshore Engineer-
nitude of force from the pre-generated database.
ing. CRC Press, Florida, USA, 2024;280. ISBN:
This force is distributed among the vessel’s four 9781032806068.
thrusters, precisely counteracting external distur- 3. Molin B. On the piston and sloshing modes in
bances and returning the ship to its intended po- moonpools. J Fluid Mech. 2001;430:27-50.
sition. The results demonstrate that this method http://dx.doi.org/10.1017/S0022112000002871
213