KT Deploys 5G and AI Welding Robots at HD Hyundai Samho Shipyard in South Korea, Accelerating Physical AI in Industrial Settings
KT, a leading telecommunications company, is set to launch a groundbreaking physical AI demonstration at HD Hyundai Samho’s shipyard in Yeongam, South Jeolla Province. This innovative project combines 5G technology with graphics processing unit (GPU)-based AI computing infrastructure to automate the control of welding robots. The main goal is to establish a network environment in an industrial setting where AI can analyze camera footage and sensor data in real time to enhance robot perception, judgment, and action.
Advancing Towards Autonomous Operations
Traditionally, welding robots at shipyards have operated in a semi-automatic mode, requiring human intervention for transportation and handling unexpected scenarios. However, with the implementation of the high-performance AI network by KT, these robots will be able to transmit video and sensor data over the network for real-time analysis, enabling autonomous task execution. The use of AI technology will empower welding robots to perform precision tasks without constant human oversight, leading to increased efficiency and productivity in the shipyard operations.
Furthermore, AI will also be integrated into the painting process, where AI painting robots will analyze hull surfaces in real time and autonomously identify areas that require coating. This shift towards automation in painting tasks not only improves workplace safety but also enhances overall productivity by streamlining the painting process.
Integrating Cutting-edge Technologies for Enhanced Performance
The high-performance AI network at the shipyard leverages 5G standalone (SA) technology, network slicing, AI radio access network (AI-RAN), GPU-based AI computing infrastructure, and autonomous network technologies. Network slicing ensures consistent communication quality for robot control, even during peak data usage periods within the shipyard.
AI-RAN technology optimizes communication and computing resources in real time based on robot work conditions, enhancing overall network and computing infrastructure efficiency. By distributing AI computing based on task characteristics, the system improves response times and resource utilization, addressing previous limitations faced by AI robots.
In parallel, SK Telecom consortium is conducting similar demonstrations at petrochemical and automotive sites to validate the applicability of physical AI in various industries facing labor shortages and industrial risks. Through these initiatives, the Ministry of Science and ICT aims to pave the way for the widespread adoption of physical AI in industrial settings.
In conclusion, these innovative projects by KT and SK Telecom mark significant advancements in the integration of AI technology into industrial operations, highlighting the potential for increased efficiency, safety, and productivity in manufacturing sectors.
