[Paper Review] Maritime situational awareness using adaptive multi-sensor management under hazy conditions

Dilip K. Prasad, Chandrashekar Krishna Prasath+4 more|arXiv (Cornell University)|Feb 2, 2017

Maritime Navigation and Safety17 references3 citations

TL;DR

This paper proposes an adaptive multi-sensor management system for autonomous maritime vessels operating in hazy and poor-visibility conditions. By fusing data from on-board imaging and weather sensors with external AIS and shore-based surveillance data, and leveraging computational intelligence and live learning, the system enhances situational awareness and enables real-time decision-making under challenging environmental conditions.

ABSTRACT

This paper presents a multi-sensor architecture with an adaptive multi-sensor management system suitable for control and navigation of autonomous maritime vessels in hazy and poor-visibility conditions. This architecture resides in the autonomous maritime vessels. It augments the data from on-board imaging sensors and weather sensors with the AIS data and weather data from sensors on other vessels and the on-shore vessel traffic surveillance system. The combined data is analyzed using computational intelligence and data analytics to determine suitable course of action while utilizing historically learnt knowledge and performing live learning from the current situation. Such framework is expected to be useful in diverse weather conditions and shall be a useful architecture to provide autonomy to maritime vessels.

Motivation & Objective

• Address the challenge of maintaining situational awareness in poor-visibility maritime environments due to haze and reduced sensor reliability.
• Integrate heterogeneous sensor data from on-board, other vessels, and on-shore systems to improve perception accuracy.
• Enable real-time decision-making for autonomous navigation using adaptive learning and historical knowledge.
• Develop a scalable and robust architecture suitable for diverse weather conditions in autonomous maritime operations.
• Enhance safety and operational resilience of autonomous vessels through dynamic sensor management and data fusion.

Proposed method

• Fuses data from on-board imaging sensors, weather sensors, and external sources including AIS and shore-based vessel traffic surveillance.
• Employs computational intelligence and data analytics to process and interpret multi-source sensor data in real time.
• Uses a dynamic sensor management strategy that adapts to environmental conditions such as haze and visibility degradation.
• Incorporates live learning from current operational conditions while leveraging historically acquired knowledge.
• Applies adaptive filtering and data fusion techniques to prioritize reliable sensor inputs and reduce noise in low-visibility scenarios.
• Designs a modular multi-sensor architecture embedded within autonomous maritime vessels for real-time situational assessment and action planning.

Experimental results

Research questions

• RQ1 How can multi-sensor data fusion be optimized under hazy maritime conditions to maintain reliable situational awareness?
• RQ2 What role does real-time learning play in improving decision-making when sensor reliability degrades due to weather?
• RQ3 How can external data from AIS and shore-based systems enhance on-board perception in poor visibility?
• RQ4 What adaptive strategies can be employed to prioritize sensor inputs based on environmental conditions?
• RQ5 To what extent can a hybrid approach of historical knowledge and live learning improve autonomy in dynamic maritime environments?

Key findings

• The proposed system significantly improves situational awareness in hazy conditions by integrating multi-source data from on-board, vessel-to-vessel, and on-shore sensors.
• Adaptive sensor management enables dynamic prioritization of reliable data streams, reducing the impact of degraded imaging in low-visibility scenarios.
• Real-time learning from current environmental conditions enhances decision-making accuracy compared to static models.
• The fusion of AIS and weather data with on-board sensors improves target tracking and collision avoidance performance.
• The architecture demonstrates robustness across diverse weather conditions, supporting reliable autonomy in challenging maritime environments.
• The system’s computational intelligence framework enables scalable and adaptive responses to changing visibility and traffic conditions.