Page 483 - AI for Good Innovate for Impact
P. 483
AI for Good Innovate for Impact
4�5� Category 5: Manufacturing
Use Case - 1: Morphing-BOT: Revolutionizing Global Infrastructure 4.5: Manufacturing
with AI-Powered Smart Inspection and Predictive Maintenance
Organization: Morphing I
Country: South Korea
Contact Person(s):
Kevin Kwon , dohoon57@ morphingi .com
Kiyoung Kim, kiyoungkim@ morphingi.
Kyeongmin Lee, km0107@ morphingi.
1 Use Case Summary Table
Item Details
Category Manufacturing: Robot and AI-assisted maintenance
Problem Aging and buried water pipelines are difficult to access and inspect, especially
Addressed in complex, curved, or narrow sections. Traditional inspection methods are
labor-intensive, invasive, and limited in real-time data acquisition. This results
in undetected anomalies, delayed maintenance, and increased risk of leaks,
bursts, and water losses in urban infrastructure.
Key Aspects of Morphing I, provides an AI-powered pipeline inspection robot, “Morph-
Solution ing-BOT,” integrated with IoT sensors and a digital twin platform.
The robot navigates narrow and curved pipes autonomously, detects anom-
alies, collects multi-modal data (acoustic, vibration, image), and transmits
real-time insights to a digital twin system for predictive maintenance.
Technology AI Robot, Smart Infrastructure, Digital Twin, Predictive Maintenance, Pipe
Keywords Inspection, Acoustic AI, Visual AI
Data Availability Private (Data collected from live testbeds in South Korea. Over 110,000
labelled data points collected and validated by TTA[1])
Metadata (Type Image, Acoustic (hydraulic noise), Lidar, Environmental (pressure, humidity)
of Data)
Model Training Supervised learning for anomaly classification using CNNs and transform-
and Fine-Tuning ers; domain-specific fine-tuning for different pipeline types and countries;
on-device lightweight models for real-time inference and off-device training
for global optimization.
447
