Digital Twin-Based Transport Robot System and Monitoring Method

Digital Twin-Based Predictive Maintenance and Monitoring System for Transport Robots

This technology creates a digital twin on a server using real-time monitoring data (audio, IMU, and component usage counts) from transport robots. It then calculates failure probability and remaining useful life through a predictive model, visualizing the results in an XR environment.

Maintenance efficiency is currently low due to the difficulty of accurately predicting the status and failure probability of transport robots in real-time, as well as a lack of simulation for actual operating environments and intuitive status visualization.

This technology builds a predictive model that converts the robot's audio data into images to extract features, then integrates these with IMU sensor values and component usage data via a Fully Connected Layer (FCL) to determine failure and remaining lifespan.

‍

‍

Key Features:

• Transport robot that moves items based on transport information—including origin, destination, and item type—while collecting audio data, IMU sensor values, and component usage counts.
• Server that generates a digital twin based on the robot's monitoring data and transport information, predicts lifespan and failure, and visualizes transport simulations.
• Monitoring terminal that receives transport simulations from the server and displays them for user verification.
• Predictive model that calculates failure probability and expected lifespan by integrating feature data (converted and smoothed audio images) with IMU sensor values and component life data using an FCL.

‍