Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing

Tien Thanh Nguyen; Maxim Vochten; Hafeez Chaudhary; Geert De Cubber

doi:10.1007/978-3-032-29469-2_27

Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing

Robotics & Autonomous Systems

Agilica BV

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Downloads (Pure)

Abstract

Autonomous landing of Unmanned Aerial Vehicles (UAVs) on maritime vessels is challenging due to stochastic wave-induced motions. This paper proposes a realtime robust deep learning framework for long-term ship motion prediction using exclusively Inertial Measurement Unit (IMU) data. To mitigate risks from sudden vessel or environmental shifts, an anomaly detection mechanism verifies data stability prior to forecasting. The system is trained via a high-fidelity simulator
and validated against real-world vessel telemetry, demonstrating strong sim-to-real generalization. Integrating this network into the UAV control loop enables the determination of optimal landing windows and precise attitude synchronization, significantly reducing landing impact forces. Code and datasets are provided for open access.

Original language	English
Title of host publication	Robotics and Mechatronics - Proceedings of ISRM 2026
Editors	Vu Linh Nguyen, Phong Dien Nguyen, Nguyen Quang Hoang, Giuseppe Carbone
Publisher	Springer Science and Business Media B.V.
Pages	259-268
Number of pages	10
ISBN (Print)	9783032294685
DOIs	https://doi.org/10.1007/978-3-032-29469-2_27 https://doi.org/10.1007/978-3-032-29469-2_27
Publication status	Published - 2026
Event	9th IFToMM International Symposium on Robotics and Mechatronics, ISRM 2026 - Hanoi, Viet Nam Duration: 15 Jul 2026 → 17 Jul 2026

Publication series

Name	Mechanisms and Machine Science
Volume	213
ISSN (Print)	2211-0984
ISSN (Electronic)	2211-0992

Conference

Conference	9th IFToMM International Symposium on Robotics and Mechatronics, ISRM 2026
Country/Territory	Viet Nam
City	Hanoi
Period	15/07/26 → 17/07/26

Keywords

anomaly detection
deep time-series forecasting
maritime UAV landing
ship motion prediction

Access to Document

Cite this

Nguyen, T. T., Vochten, M., Chaudhary, H., & De Cubber, G. (2026). Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing. In V. L. Nguyen, P. D. Nguyen, N. Quang Hoang, & G. Carbone (Eds.), Robotics and Mechatronics - Proceedings of ISRM 2026 (pp. 259-268). (Mechanisms and Machine Science; Vol. 213). Springer Science and Business Media B.V.. https://doi.org/10.1007/978-3-032-29469-2_27, https://doi.org/10.1007/978-3-032-29469-2_27

Nguyen, Tien Thanh ; Vochten, Maxim ; Chaudhary, Hafeez et al. / Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing. Robotics and Mechatronics - Proceedings of ISRM 2026. editor / Vu Linh Nguyen ; Phong Dien Nguyen ; Nguyen Quang Hoang ; Giuseppe Carbone. Springer Science and Business Media B.V., 2026. pp. 259-268 (Mechanisms and Machine Science).

@inproceedings{24a24d86614548ce8aaa50d960048cf6,

title = "Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing",

abstract = "Autonomous landing of Unmanned Aerial Vehicles (UAVs) on maritime vessels is challenging due to stochastic wave-induced motions. This paper proposes a realtime robust deep learning framework for long-term ship motion prediction using exclusively Inertial Measurement Unit (IMU) data. To mitigate risks from sudden vessel or environmental shifts, an anomaly detection mechanism verifies data stability prior to forecasting. The system is trained via a high-fidelity simulator and validated against real-world vessel telemetry, demonstrating strong sim-to-real generalization. Integrating this network into the UAV control loop enables the determination of optimal landing windows and precise attitude synchronization, significantly reducing landing impact forces. Code and datasets are provided for open access.",

keywords = "anomaly detection, deep time-series forecasting, maritime UAV landing, ship motion prediction",

author = "Nguyen, \{Tien Thanh\} and Maxim Vochten and Hafeez Chaudhary and \{De Cubber\}, Geert",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2027.; 9th IFToMM International Symposium on Robotics and Mechatronics, ISRM 2026 ; Conference date: 15-07-2026 Through 17-07-2026",

year = "2026",

doi = "10.1007/978-3-032-29469-2\_27",

language = "English",

isbn = "9783032294685",

series = "Mechanisms and Machine Science",

publisher = "Springer Science and Business Media B.V.",

pages = "259--268",

editor = "Nguyen, \{Vu Linh\} and Nguyen, \{Phong Dien\} and \{Quang Hoang\}, Nguyen and Giuseppe Carbone",

booktitle = "Robotics and Mechatronics - Proceedings of ISRM 2026",

}

Nguyen, TT , Vochten, M, Chaudhary, H & De Cubber, G 2026, Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing. in VL Nguyen, PD Nguyen, N Quang Hoang & G Carbone (eds), Robotics and Mechatronics - Proceedings of ISRM 2026. Mechanisms and Machine Science, vol. 213, Springer Science and Business Media B.V., pp. 259-268, 9th IFToMM International Symposium on Robotics and Mechatronics, ISRM 2026, Hanoi, Viet Nam, 15/07/26. https://doi.org/10.1007/978-3-032-29469-2_27, https://doi.org/10.1007/978-3-032-29469-2_27

Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing. / Nguyen, Tien Thanh ; Vochten, Maxim; Chaudhary, Hafeez et al.
Robotics and Mechatronics - Proceedings of ISRM 2026. ed. / Vu Linh Nguyen; Phong Dien Nguyen; Nguyen Quang Hoang; Giuseppe Carbone. Springer Science and Business Media B.V., 2026. p. 259-268 (Mechanisms and Machine Science; Vol. 213).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing

AU - Nguyen, Tien Thanh

AU - Vochten, Maxim

AU - Chaudhary, Hafeez

AU - De Cubber, Geert

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2027.

PY - 2026

Y1 - 2026

N2 - Autonomous landing of Unmanned Aerial Vehicles (UAVs) on maritime vessels is challenging due to stochastic wave-induced motions. This paper proposes a realtime robust deep learning framework for long-term ship motion prediction using exclusively Inertial Measurement Unit (IMU) data. To mitigate risks from sudden vessel or environmental shifts, an anomaly detection mechanism verifies data stability prior to forecasting. The system is trained via a high-fidelity simulator and validated against real-world vessel telemetry, demonstrating strong sim-to-real generalization. Integrating this network into the UAV control loop enables the determination of optimal landing windows and precise attitude synchronization, significantly reducing landing impact forces. Code and datasets are provided for open access.

AB - Autonomous landing of Unmanned Aerial Vehicles (UAVs) on maritime vessels is challenging due to stochastic wave-induced motions. This paper proposes a realtime robust deep learning framework for long-term ship motion prediction using exclusively Inertial Measurement Unit (IMU) data. To mitigate risks from sudden vessel or environmental shifts, an anomaly detection mechanism verifies data stability prior to forecasting. The system is trained via a high-fidelity simulator and validated against real-world vessel telemetry, demonstrating strong sim-to-real generalization. Integrating this network into the UAV control loop enables the determination of optimal landing windows and precise attitude synchronization, significantly reducing landing impact forces. Code and datasets are provided for open access.

KW - anomaly detection

KW - deep time-series forecasting

KW - maritime UAV landing

KW - ship motion prediction

UR - https://www.scopus.com/pages/publications/105045216175

U2 - 10.1007/978-3-032-29469-2_27

DO - 10.1007/978-3-032-29469-2_27

M3 - Conference contribution

AN - SCOPUS:105045216175

SN - 9783032294685

T3 - Mechanisms and Machine Science

SP - 259

EP - 268

BT - Robotics and Mechatronics - Proceedings of ISRM 2026

A2 - Nguyen, Vu Linh

A2 - Nguyen, Phong Dien

A2 - Quang Hoang, Nguyen

A2 - Carbone, Giuseppe

PB - Springer Science and Business Media B.V.

T2 - 9th IFToMM International Symposium on Robotics and Mechatronics, ISRM 2026

Y2 - 15 July 2026 through 17 July 2026

ER -

Nguyen TT , Vochten M, Chaudhary H, De Cubber G. Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing. In Nguyen VL, Nguyen PD, Quang Hoang N, Carbone G, editors, Robotics and Mechatronics - Proceedings of ISRM 2026. Springer Science and Business Media B.V. 2026. p. 259-268. (Mechanisms and Machine Science). doi: 10.1007/978-3-032-29469-2_27, 10.1007/978-3-032-29469-2_27

Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

SAILS: Autonomous Maritime UAS Landing Solution

Ship Motion Prediction

Cite this

Deep Long-Term Ship Motion Prediction with Anomaly Detection for Autonomous Maritime UAV Landing

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Projects

SAILS: Autonomous Maritime UAS Landing Solution

Datasets

Ship Motion Prediction

Cite this