OKI develops AI-driven technology for ship classification using underwater sounds

OKI, a Tokyo-based information and telecommunications technology provider, has unveiled an advanced ship classification AI system that leverages underwater sounds to classify ship types with remarkable precision. This state-of-the-art technology is poised to redefine maritime monitoring, achieving classification accuracy of over 90%, even with minimal learning data.

By combining deep learning with underwater sound analysis, OKI’s innovation offers an automated solution to ship classification challenges, particularly in high-traffic maritime environments or during nighttime when visual identification using cameras is ineffective. This breakthrough builds on OKI’s extensive expertise in underwater acoustics and aligns with the increasing demand for labor-saving solutions in the maritime industry.

How Does OKI’s AI Classify Ships Using Underwater Sounds?

Unlike traditional methods that rely on radio or light waves, OKI’s system harnesses the unique properties of sound waves, which can travel underwater for thousands of kilometers. Each ship emits distinct sound frequencies, allowing its type to be identified. OKI’s AI system processes these frequencies using hydrophones (underwater microphones) to create deep learning models. These models are then used to automatically classify ships, eliminating the need for human intervention.

This system addresses significant challenges posed by traditional methods, such as relying on human expertise to interpret sound data. Human classification often leads to inconsistencies due to skill variability and fatigue. OKI’s automated approach ensures consistent, reliable results while reducing human labor requirements.

What Challenges Does the AI System Overcome?

The maritime industry has long relied on sound wave analysis for underwater classification, as other methods face severe limitations. Radio waves are significantly attenuated underwater, and light waves are widely scattered, making them unreliable. While human operators have traditionally analyzed sound frequencies, this process is not only labor-intensive but also prone to error.

OKI’s ship classification AI system addresses these issues through advanced algorithms that analyze the characteristics of ship sounds. The system trains its deep learning models using underwater sound data, enabling it to classify ships without depending on human skill levels. Additionally, the AI system uses innovative data augmentation techniques to artificially create variations in existing sound data, expanding its training dataset without requiring extensive new recordings.

Achieving High Accuracy with Limited Data

A major highlight of OKI’s AI system is its ability to achieve over 90% accuracy despite being trained on just four hours of ship sound data. Typically, deep learning models require vast amounts of training data to perform effectively. However, acquiring and preparing such datasets in the maritime sector poses significant challenges due to time, cost, and logistical constraints.

To overcome these limitations, OKI employed semi-supervised learning techniques that allow the AI to extract meaningful insights even from incomplete datasets. By combining semi-supervised learning with data augmentation, OKI’s system optimizes its deep learning model to deliver high accuracy while minimizing data requirements.

This advancement not only reduces the cost and effort of dataset preparation but also demonstrates the potential for scalability in commercial applications.

Why This Innovation Matters for the Maritime Industry

OKI’s ship classification AI system holds immense potential for the maritime sector. With global trade relying heavily on shipping, efficient monitoring and classification of vessels are critical for port operations, maritime security, and environmental management.

The system’s ability to operate continuously and autonomously, even in high-traffic areas or at night, provides a significant edge over traditional methods. Its labor-saving capabilities also align with the growing demand for automation across industries.

According to Yoichi Kato, Senior Executive Officer and Head of the TOKKI Systems Division at OKI, the company is actively seeking co-creation partners to gather field data and conduct further practical verifications. These efforts aim to pave the way for the commercialization of this transformative technology.

The Road Ahead: Expanding AI’s Role in Maritime Innovation

As OKI works toward commercializing its ship classification AI system, its broader implications for the maritime industry are becoming increasingly evident. From improving port operations to enhancing environmental monitoring, the system has the potential to revolutionize how ships are classified and tracked.

By leveraging its decades-long expertise in underwater acoustic research, OKI continues to push the boundaries of innovation. This latest development reinforces its position as a global leader in information and telecommunications technology, dedicated to delivering cutting-edge solutions that address real-world challenges.