Abstract:The clearance of explosive remnants of war (ERW) continues to be a predominantly manual and high-risk process that can benefit from advances in technology to improve its efficiency and effectiveness. In particular, research on artificial intelligence for ERW clearance has grown significantly in recent years. However, this research spans a wide range of fields, making it difficult to gain a comprehensive understanding of current trends and developments. Therefore, this article provides a literature review of academic research on AI for ERW detection for clearance operations. It finds that research can be grouped into two main streams, AI for ERW object detection and AI for ERW risk prediction, with the latter being much less studied than the former. From the analysis of the eligible literature, we develop three opportunities for future research, including a call for renewed efforts in the use of AI for ERW risk prediction, the combination of different AI systems and data sources, and novel approaches to improve ERW risk prediction performance, such as pattern-based prediction. Finally, we provide a perspective on the future of AI for ERW clearance. We emphasize the role of traditional machine learning for this task, the need to dynamically incorporate expert knowledge into the models, and the importance of effectively integrating AI systems with real-world operations.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: How to use artificial intelligence (AI) technology to improve the efficiency and effectiveness of explosive remnants of war (ERW) clearance operations. Specifically, the paper aims to review existing academic research, evaluate the current application of AI in ERW detection, and identify opportunities for future research.
### Problem Background
The clearance of explosive remnants of war (ERW) remains a largely manual and high - risk process, especially in post - war areas. With the progress of technology, especially the development of artificial intelligence, researchers hope to introduce AI technology to improve the efficiency and safety of this process. However, current research involves multiple fields, including electrical engineering, statistics, etc., which makes it difficult to fully understand current research trends and developments.
### Research Objectives
To meet this challenge, the paper has carried out the following tasks:
1. **Literature Review**: A systematic review of 100 research papers on AI systems for ERW clearance published between 1997 and 2024 was conducted.
2. **Classification Analysis**: The research was divided into two mainstreams:
- **ERW Object Detection**: Use AI technology to detect ERW objects from sensor data or image data.
- **ERW Risk Prediction**: Predict the risk of ERW in a certain area based on historical data and other relevant factors.
3. **Identify Research Gaps**: Analyze the shortcomings of existing research and propose directions for future research.
### Main Findings
- **ERW Object Detection**: More than 90% of the research focuses on using sensor data (such as ground - penetrating radar (GPR), metal detectors, etc.) and image data (such as thermal imaging, hyperspectral imaging, etc.) to detect ERW objects. Commonly used AI models include convolutional neural networks (CNN), support vector machines (SVM), etc.
- **ERW Risk Prediction**: Fewer studies focus on risk prediction, but these studies combine ERW - related data (such as accident records) and non - ERW - related data (such as terrain, infrastructure information) to conduct risk assessment.
### Future Research Directions
The paper proposes three opportunities for future research:
1. **Strengthen ERW Risk Prediction Research**: Further explore and improve risk prediction models, for example, by identifying more effective input features and comparing the performance of different machine - learning techniques.
2. **Combine Different AI Systems**: Explore how to combine different AI systems to improve the overall detection effect, for example, by fusing image data with sensor data.
3. **Introduce New Models and Methods**: Develop new algorithms and techniques, such as pattern - based prediction methods, to improve the accuracy of ERW risk prediction.
Through these efforts, researchers hope that AI can become a powerful tool for improving the efficiency and safety of ERW clearance operations.