Jakarta, INTI - There are many damaged roads across various regions in Indonesia, ranging from rural villages and cities to provincial areas. However, road condition monitoring in Indonesia has so far still been carried out manually, making the process time-consuming. So, what is the solution?
To address this issue, a doctoral researcher from the Institut Teknologi Sepuluh Nopember (ITS), Hani’ah Mahmudah, has developed an AI-based road damage detection system. The tool utilizes a Convolutional Neural Network (CNN) model running on edge computing devices.
With this system, road damage can be automatically identified through camera sensors installed on vehicles.
“The detection optimization used is based on artificial intelligence (AI) to speed up real-time data collection accurately,” she said, as quoted from the official ITS website on Wednesday, April 15, 2026.
How the AI Road Damage Detection Works and Its Benefits
The system works by collecting visual data in the form of images or videos of road conditions. This data is then analyzed using an AI model to identify damage such as potholes or cracks.
Hani’ah explained that the visual data is processed using the iYOLOv7-TPE-SS model to achieve more accurate object detection. The results are then processed using AI-based computing hardware, namely NVIDIA AGX Orin.
From this process, the system produces road condition information that can be directly used for maintenance planning. This technology is considered capable of significantly improving data collection efficiency compared to conventional methods.
She noted that this innovation is not only beneficial for academic purposes but also has broad potential for real-world application. Automated detection systems can help governments reduce road maintenance costs while improving user safety.
“Moreover, faster and more objective road inspections using AI also support the development of smart cities and the Internet of Vehicles (IoV),” she explained.
Research Published in International Journals
During her studies, Hani’ah also produced several scientific publications in international journals. Seven out of her nine research papers were published by the Institute of Electrical and Electronics Engineers (IEEE).
Her research focuses on machine learning and digital signal processing. Hani’ah hopes that her innovation will attract industry partners so that the technology can be widely implemented.
This development also supports ITS’s commitment to achieving the Sustainable Development Goals (SDGs), particularly in quality education, industry innovation and infrastructure, and sustainable cities.
Conclusion
AI-based road damage detection developed by ITS offers a faster, more accurate, and scalable solution for infrastructure monitoring in Indonesia. By automating road inspections through edge computing and deep learning, this innovation can improve maintenance efficiency, enhance road safety, and support the development of smart cities and sustainable transportation systems.
Read more: Indonesia Affirms Active Role in Global AI Governance at the IMF World Bank Spring Meetings 2026