event-trade登录编辑
Seoul Economic Daily EnglishCodex · 实盘不做把握 7/10
北京时间·🕒 发布 14:00:05时区推断·⟳ 抓取 14:32:33·⚡ 决策 07-14 14:50:03 +17m30s·105.2 天前

SK Telecom与KT入选韩国Hyper AI Network试点并获得172.06亿韩元两年投资

项目两年总投资172.06亿韩元,由SKT和KT两个联合体分摊,对大型电信运营商收入和市值量级很小。

📄 原文Seoul Economic Daily English↗ 原文
SKT, KT Launch AI-RAN Project at Core of Korea's 'AI Highway'
SK Telecom (017670.KS) and KT (030200.KS) will build an "AI-RAN (AI-based radio access network) pilot network," a core…
不做 SK Telecom(017670.KS)
两年172.06亿韩元项目对大型运营商量级太小,且更偏试点验证,缺少短线重定价空间,不做。
不做 KT Corp(030200.KS)
项目金额和商业化节奏不足以改变KT短期收入利润,且设备份额不明,不做。
展开详情(成交明细 / 研究档案 / 原文全文 / 决策记录链)

研究档案

🔎 新鲜度:最早出现 2026-07-14 · 扩散 medium · price-in 风险 medium ↗

消息传播链

  1. 2026-07-14官方确认/否认MSIT/NIA cited by Seoul Economic Daily一手源原文 ↗
    韩国科学技术信息通信部和NIA宣布SKT、KT入选试点项目,两年总投资172.06亿韩元。
  2. 2026-07-14行业媒体转载Seoul Economic Daily非一手原文 ↗
    报道披露SKT将测试三星、HFR、爱立信、诺基亚设备,KT将联合HD Hyundai Samho测试船厂物理AI。

关键理由

  • 项目两年总投资172.06亿韩元,由SKT和KT两个联合体分摊,对大型电信运营商收入和市值量级很小。
  • 报道列出三星电子、HFR、爱立信、诺基亚等设备参与方,但没有披露各公司订单金额和毛利贡献。

主要风险

  • HFR等小型通信设备商若后续披露实际中标金额,可能具备更高弹性。
  • 国家AI-RAN示范若演化为大规模6G基础设施投资,需由新政策或订单信号重新评估。

反向证据

  • 做多前查反向 · 来源 web · 日期 2026-07-14 · 影响 weakens
    项目金额分散且供应商份额不明,对SKT、KT、三星等上市公司均缺少可量化短期业绩弹性。
    检索:SKT KT AI-RAN pilot network 17.206 billion won equipment supplier amount 原文 ↗
工具调用 · 历史记录,结果未知(1 种)
  • normalize_ticker · 结果未知

原文全文

展开原文 · 7,127 字
SK Telecom (017670.KS) and KT (030200.KS) will build an "AI-RAN (AI-based radio access network) pilot network," a core element of the "artificial intelligence (AI) highway" national initiative, and pursue the development and testing of services converged with physical AI.
The Ministry of Science and ICT (MSIT) and the National Information Society Agency (NIA) announced on the 14th that SKT and KT were selected as the executing organizations for the "Hyper AI Network Foundation" pilot project. A total of 17.206 billion won will be invested in the two consortiums over two years.
The "Hyper AI Network" is a next-generation network that applies AI to telecommunications networks to provide the ultra-low-latency, high-capacity communications needed for physical AI services. The key is for the network to share the computation load of physical AI, reducing the burden on devices while improving performance and efficiency. This will enable physical AI services such as welding, painting, and patrol robots to make agile, real-time judgments and movements at industrial sites.
In this process, KT will build an AI-based autonomous operation network and verify a communications environment in which multiple robots can operate stably at industrial sites. SKT plans to apply 5G SA-specific technologies such as AI-RAN and network slicing, along with an integrated management system (SMO) and AI-based autonomation technologies to the pilot network, quantitatively verifying performance improvement effects.
SKT to Build, Test Pilot Network with Equipment from Four Firms Including Samsung Electronics, HFR, Ericsson, Nokia
In this project, SKT will simultaneously develop, build, and test AI-RAN equipment from four manufacturers: Samsung Electronics, HFR, Ericsson, and Nokia.
SKT will also apply various computing resources such as CPUs and GPUs to compare performance in the same environment, and will operate the AI server and data transfer device (UPF) in various configurations to verify the structure best suited for physical AI services.
KT to Test Three Types of Shipyard Physical AI with Samsung Electronics, HD Hyundai Samho
KT was tasked with establishing an AI-based autonomous operation network and verifying a communications environment in which multiple robots can operate stably at industrial sites such as shipyards.
In particular, as the foundation of the autonomous operation network, KT will develop an "AI Core Orchestrator" in which AI analyzes network data in real time and addresses failures. To this end, KT plans to link the "Network Data Analytics Function (NWDAF)" with AI to train it on the communication patterns and performance data of the core network.
In cooperation with HD Hyundai Samho, KT will also develop an "AI-based autonomous system" specialized for shipyards. Based on the ultra-low-latency communications environment of the Hyper AI Network, AI will be able to analyze and control shipyard robot and equipment data in real time. Specifically, KT plans to test a total of three types of physical AI services—an AI welding robot, an AI painting robot, and an autonomous operation robot for communications facilities—to verify the effects of improving safety and productivity at high-risk industrial sites.
In addition, KT will pursue a "K-telecommunications ecosystem revitalization" task. KT will build a testbed for verifying domestic equipment at its KT Research and Development Center in Umyeon-dong to develop base station power reduction technology and low-power 5G (RedCap) technology, and will support the overseas expansion of the domestic telecommunications ecosystem and companies.
Three Types of Physical AI Services Tested: Patrol Robots, Humanoids, Autonomous Transport
SKT will also test three types of physical AI services based on the AI-RAN pilot network: patrol robots, humanoids, and unmanned autonomous transport services. Through this, SKT plans to quantitatively verify high-capacity uplink transmission, ultra-low latency, distributed computation, mobility, and reliability.
The quadruped patrol robot will be tasked with transmitting high-definition video in real time without interruption while patrolling hazardous areas within a factory. AI-RAN will analyze this video with an AI model to provide a service that performs everything from hazard detection to integrated control.
The unmanned autonomous transport service gathers LiDAR sensor data installed in on-site infrastructure to the AI-RAN base station via ultra-low-latency, high-reliability communications. The base station centrally processes this data, composes a digital twin that reproduces the site as-is, and then issues remote driving commands to vehicles. The goal is to automate factory logistics, thereby improving efficiency and lowering operating costs.
The humanoid low-power mode is a service that distributes the robot's complex AI computations to the AI-RAN base station, reducing the device's computation burden and battery consumption. This addresses the problem of short device battery life and enables similar performance to be achieved even with cheaper robots.
In the first year of the project, SKT will build AI-RAN pilot networks in Incheon and Pangyo and test two types of physical AI services. At SK Incheon Petrochem in Incheon, SKT plans to verify an industrial safety control service using quadruped patrol robots and mobile CCTV, while at the Pangyo Physical AI Living Lab, it will verify an unmanned autonomous transport service.
In the second year, the testing will expand to actual industrial sites. The industrial safety control service in Incheon will be advanced, and the unmanned autonomous transport service verified in Pangyo will be applied to KG Mobility's Pyeongtaek plant. Alongside this, SKT will also verify AI-based network autonomous control and low-latency, high-reliability communications technology.
AI-RAN Project Linked to Operators' Next-Generation Growth Engines
KT and SKT plan to link the technology secured through this project with their respective AX strategies.
Lee Jong-sik, executive vice president and head of KT's Future Network Lab, stressed, "Based on our top-tier network operation capabilities in Korea, we will discover core technologies for the 6G era," adding, "We will lead the expansion of the Hyper AI ecosystem and contribute to enhancing the competitiveness of the national telecommunications equipment industry." KT plans to expand into various industries such as manufacturing, logistics, and energy to support companies' AX transformation.
SKT also announced that it will link the AI-RAN pilot network with SK Group's AI data center infrastructure strategy going forward. Ryu Tak-gi, head of SKT's network technology, said, "As the only Korean member of the AI-RAN Alliance board of directors, we will test various AI-RAN pilot networks and physical AI services based on the capabilities we have accumulated." He added, "While advancing AI-RAN technology, we will raise both the self-reliance and global competitiveness of the domestic ecosystem through win-win cooperation among large, mid-sized, and small companies."

决策记录链 · 只追加

每次领取产生一条记录,重新决策追加一条并以 supersedes 指向被替代的记录(ADR 0014)。

这条决策早于决策账本(2026-10-08 起逐条追加),没有追加记录。