border border


Team members

Lee Liang Heng Caleb (ISTD), Ang Wei Jie (EPD), Leong Kei Sheng (EPD), Vieri Vincent (ISTD), Louth Bin Rawshan (EPD), Sarah Ong (ESD), Lee Rui Lin Eunice (ESD)


Yeo Si Yong, Dawn Tan, Georgios Piliouras

Writing Instructors:

Pang Yoke Kian Rachel

Teaching Assistant:

Eve Cheong Yew Kuan

Problem Statement

Drones have revolutionized tank inspection by replacing hazardous scaffolding and human scaling. However, the post-inspection process, from reviewing footage to the final report, is labor intensive and time-consuming. The manual identification and location of defects from drone footage, as well as report generation, have been identified as the bottlenecks. An end-to-end automation is necessary.


Locating and identifying structural defects automatically for drone based inspection to be scalable






Identifying Defects Automatically

Using the latest deep learning techniques,  our model is capable of detecting 4 critical steel defects - Pitting, Cracks, Stains, Corrosion and achieves state of the art performance of ~70% accuracy.



Modular Payload





All-in-One Cloud Hosted Web Platform

An all in one cloud hosted web platform for you to visualize and verify your inspection results, fulfilling all your post inspection needs.





Automated Defect Localization

Each frame from the video is mapped to its location in the floorplan automatically. This is achieved using computer vision techniques, which calculate the distance travelled by the drone by analyzing the video footage.








Fully Customized Modular Payload

A customised modular payload that provides image stabilisation and on-site defect identification tools, as well as performing sensor fusion tailored to the needs of our automated identification and localisation systems.



landing page

Industry Partner

new pr logo 2017 400x128


student Lee Liang Heng Caleb Information Systems Technology and Design
student Ang Wei Jie Engineering Product Development
student Leong Kei Sheng Engineering Product Development
student Vieri Vincent Information Systems Technology and Design
student Louth Bin Rawshan Engineering Product Development
student Sarah Ong Engineering Systems and Design
student Lee Rui Lin Eunice Engineering Systems and Design
border border