[HCI Opinion] "Roadmap for Industrial Safety and Logistics Innovation Based on the Ulsan AI Data Center"

HOSOON CHOI
Jun 28
16 min read

Ulsan’s Transformation: From Manufacturing Capital to a Smart Innovation City Powered by AI

Published on : June 25, 2025

Author Hosoon Choi, Logistics Strategy Specialist

Certified Logistics Manager | Licensed Customs Broker | PMP® | MBA

"Empowering Logistics Through Data Intelligence" - Insight from Korea's Strategic Logistics Frontline

This image was generated by OpenAI’s DALL·E

Executive Summary

Ulsan Metropolitan City has recently secured Korea’s largest AI Data Center through a strategic partnership between Amazon Web Services (AWS) and the SK Group. This major achievement establishes Ulsan as a new growth hub in the era of digital transformation. This policy proposal outlines a strategic direction that goes beyond the mere acquisition of digital infrastructure. It presents a forward-looking roadmap for Ulsan to lead AI-driven innovation across key sectors, including industrial safety, port logistics, and energy. Although Ulsan is already pursuing government-level initiatives such as smart port development and industrial complex upgrades, there remain critical policy gaps in the following convergence areas: -AI-based prediction of major industrial accidents (Safety AI) -Integrated port-industrial logistics platforms -Enterprise-level GPU infrastructure for AI solution testing -RE100-compliant, low-carbon logistics clusters To address these gaps, this proposal puts forth four actionable strategies designed to shape Ulsan’s differentiated and future-oriented vision.

1. Ulsan’s Vision Powered by Large-Scale AI Infrastructure

Ulsan Metropolitan City has recently secured Korea’s largest artificial intelligence (AI) data center, positioning itself at the forefront of industrial transformation. Backed by a massive ₩7 trillion (approx. $5 billion) investment from SK Group and Amazon Web Services (AWS), the facility will be equipped with 60,000 neural processing units (NPUs) dedicated to AI operations. The data center is set to expand in phases to a full-scale 1 GW-class complex, ultimately becoming the largest AI hub in Northeast Asia.

SK Group Chairman Chey Tae-won emphasized the project’s national significance, stating, “Ulsan was once Korea’s manufacturing stronghold. Now, it will serve as the powerful new engine of the AI superhighway.” The deployment of such a large-scale AI infrastructure is expected to spark widespread digital transformation and innovation across Ulsan’s industrial landscape.

A key factor behind Ulsan’s selection as the site for this flagship project lies in its robust energy infrastructure. A nearby LNG combined heat and power (CHP) plant ensures a stable supply of electricity, while the ultra-low temperature cold energy generated during LNG regasification (−162°C) provides a natural cooling solution for the data center’s high-performance servers. In addition, under the newly enacted Special Act on Distributed Energy Activation, Ulsan is pursuing a designation as a "Distributed Energy Zone", enabling direct power transactions between local energy producers and the data center—contributing to both cost savings and power stability.

This AI data center, seamlessly integrated with the region’s energy systems, is a cornerstone of Ulsan’s broader vision to become a smart industrial city.

This policy proposal outlines a strategic vision for leveraging the AI data center as a springboard to drive AI-powered innovation across Ulsan’s core sectors: industrial safety, port logistics, and energy.

While national and municipal-level smart port and AI demonstration projects are already underway, there remain significant policy gaps in the following four key areas:

Advanced industrial safety through AI prediction and prevention
Integrated data platforms connecting port and industrial complexes
Open access to AI infrastructure (e.g., GPU clusters) for regional SMEs
Cluster-based integration of eco-friendly data centers and green logistics hubs

To move beyond infrastructure deployment and position Ulsan as a central hub for AI-driven industrial safety and logistics transformation, this proposal presents four actionable policy initiatives in the following chapters.

2. Advanced AI for On-Site Industrial Safety and Demonstration of Major Accident Prevention

2.1 Deep Learning–Based Risk Behavior Recognition and Environmental Sensing Solutions

Even in the third year since the enforcement of the Serious Accidents Punishment Act, industrial sites in Ulsan continue to experience frequent accidents. As a result, companies are making every effort to introduce safety management solutions that incorporate AI and IoT technologies. While the focus has traditionally been on expanding CCTV coverage and on-site monitoring, there is now a need to take a step further by building systems that can predict accidents and detect early warning signs using deep learning–based vision AI and IoT sensors.

For example, the latest AI safety solutions can analyze a worker’s behavior in video footage in real time to immediately detect hazardous actions such as falls or failure to wear protective equipment, as well as anomalies like smoke or sparks, and issue alerts. In fact, a chemical plant recently adopted a system where AI sensors were attached to CCTV cameras. The system detects safety violations such as the risk of vehicle collisions or signs of gas leaks and sends instant alerts to the central control room for immediate response.

In addition, AI vision models are being used to detect environmental risk factors, such as identifying smoke before a fire occurs or recognizing structural abnormalities at an early stage. The goal is to prevent major industrial accidents through integrated monitoring of both behavior and environmental conditions.

Ulsan City can implement pilot projects applying such AI technologies to inspect the risk levels of aging facilities in industrial complexes and to monitor compliance with workplace safety rules. Especially for small and medium-sized businesses, which often lack dedicated safety personnel or the capacity to adopt their own solutions, the city should consider building a cloud-based AI safety platform that these companies can easily access and utilize.

For instance, a platform called “Ulsan AI Safety Keeper” could collect CCTV footage and IoT sensor data, allowing AI to identify dangerous behavior in real time and issue warnings. At the same time, the AI models should be continuously trained with field data to improve accuracy, and a database of industrial accidents in Ulsan should be established and used to refine the algorithms. Furthermore, from a predictive maintenance (PdM) perspective, a system that uses machine learning to analyze IoT data—such as equipment vibration and temperature—can issue early warnings of possible failures. By combining monitoring of worker behavior with equipment condition analysis, a multidimensional safety system can be established.

2.2 Establishment of Demonstration Testbeds and Regulatory Framework Alignment

Ulsan City proposes the establishment of a dedicated “AI Safety Demonstration Zone” to systematically develop and validate technologies that align with Korea’s Occupational Safety and Health Act and the Serious Accidents Punishment Act. With the headquarters of the Korea Occupational Safety and Health Agency (KOSHA) and multiple safety management institutions already located in Ulsan, the city offers an ideal environment for collaboration between industry, academia, research institutes, and government.

Based on this ecosystem, designated areas—such as sections of the Onsan National Industrial Complex—could be established as special industrial safety demonstration zones. By applying regulatory sandbox frameworks, the city can accelerate the field testing of cutting-edge safety technologies. Examples include autonomous patrol robots, smart protective gear with wearable sensors, and AR/VR-based safety training systems. Test results from these real-world trials could then inform national legal and policy improvements.

Companies participating in this initiative would gain access to a live testbed for deploying and evaluating next-generation safety technologies. Ulsan, in turn, could develop official standardized safety guidelines based on these trials, which could support regulatory refinement or the introduction of incentive mechanisms for adopting certified technologies.

In particular, Ulsan could focus on demonstration projects targeting the prevention of serious industrial accidents, such as automation of high-risk operations and technologies designed to enhance emergency response speed. For example, AI-driven alerts that signal an imminent accident could be tested alongside rapid-response shutdown protocols. The effectiveness of such systems could serve as the basis for negotiating legal incentives—such as reduced liability under the Serious Accidents Punishment Act.

Furthermore, the city could advocate, using real demonstration data, for government-backed incentives such as industrial insurance premium discounts or favorable ESG ratings for companies that adopt AI-based safety systems. By operating a shared testbed supported by both public policy and private-sector participation, Ulsan has the opportunity to position itself as the safest industrial city in Korea and a national leader in smart safety technologies.

3. Integration of Port–Industrial Complex Logistics Through a Unified Data Platform

3.1 The Need to Align with the Achievements of Smart Port Development

Ulsan, while home to one of the world’s leading industrial ports, has long faced inefficiencies due to the lack of integration between port operations data and inland logistics data. Fortunately, the Ministry of Oceans and Fisheries and the Ulsan Port Authority addressed this issue through the development of IoT-based Intelligent Port Logistics Technologies (IPLT) from 2019 to 2022, resulting in the establishment of an integrated data platform for Ulsan Port.

Building on this achievement, Ulsan recently launched PortWise, a smart port system that consolidates various data sources including the Automatic Identification System (AIS), port IoT sensors, and logistics information systems. The platform provides real-time information on vessel locations, arrivals and departures, berthing schedules, weather, and water depth conditions.

Powered by AI, PortWise also enables advanced features such as berth operation efficiency analysis, hazardous materials monitoring, and automated assessments of work productivity. These capabilities have already produced tangible benefits, including reduced vessel waiting times and improved berthing efficiency.

Now is the time to expand the success of this port digitalization effort by linking it with Ulsan’s inland industrial zones—such as the Onsan National Industrial Complex—to achieve door-to-door logistics optimization across the full supply chain.

3.2 Achieving End-to-End Logistics Visibility Between Ulsan Port and the Onsan Industrial Complex

The policy proposal aims to integrate logistics data from Ulsan Port and surrounding industrial complexes to achieve end-to-end visibility across the entire logistics cycle. To this end, the first step is to establish a port–industrial logistics integration platform.

This can be accomplished by expanding the existing PortWise system operated by the Ulsan Port Authority to include logistics information modules for industrial complexes. This would allow for seamless data flow across all stages—from port unloading to inland transportation, warehouse storage, and factory arrival.

Specifically, by linking vessel arrival forecasts (AIS data) with inland transport operations (trucks, rail), logistics schedules can be coordinated in advance. Additionally, by integrating production and shipping schedules of factories within the industrial complexes with real-time inventory data from container yards and warehouses, a Just-In-Time (JIT) logistics model can be realized.

At present, port cargo volume statistics and inland freight flow data are fragmented. However, when integrated with the national Origin-Destination tracking platform (scheduled for nationwide implementation by 2027), it will be possible to visualize the entire logistics flow—from a factory in Ulsan to an overseas market—through Ulsan Port.

This data integration will enable early identification of logistics bottlenecks, improve route optimization, and facilitate the development of alternative transport pathways.

The implementation of a unified port–industrial logistics platform would generate multiple synergistic benefits:

Cost Reduction and Efficiency GainsAI can be used to optimize port unloading equipment operations, minimize vehicle waiting times, and improve warehouse turnover rates. This would result in both time and cost savings in transportation.
Enhanced Safety and Environmental ImprovementsBy linking real-time vehicle routing data with factory hazardous materials transport information, it becomes possible to track the movement of dangerous goods and improve emergency response. For instance, Ulsan is currently developing an underground pipeline monitoring system for chemical flows; by also integrating data on tanker truck movements, the city can ensure comprehensive safety management from sea to land. Reduced wait times also mean fewer vehicle emissions, contributing to environmental sustainability.
Logistics Service InnovationCompanies in industrial complexes would be able to track cargo location and receive real-time Estimated Time of Arrival (ETA) alerts, enabling them to flexibly adjust production schedules. New services could also emerge—such as shared cargo deliveries between port and complex, and unmanned night-time freight handovers.

From a policy perspective, data standardization and openness are prerequisites for successful implementation. There is a need to develop common standards and secure integration frameworks for linking customs Port-MIS data, terminal operation system (TOS) data, and corporate ERP/logistics systems.

Ulsan City should take the lead in forming a working group with relevant agencies to drive the data standardization process. It should also provide incentives to participating companies that contribute their data.

To finance platform development, the city can actively pursue funding through national smart logistics grant programs. If needed, pilot projects can be initiated using Ulsan’s own budget. The platform could be co-managed by the Ulsan Port Authority (UPA) and relevant agencies in charge of the Onsan Industrial Complex. In the long term, this system could be expanded to include neighboring logistics hubs such as Busan and Gyeongnam, forming a wider Southeastern Korea smart logistics network.

4. Shared Use of AI Data Center GPU Clusters – Support for Local Enterprises

4.1 Opening Up Computing Resources to Accelerate AI Transformation in Local Industries

The large-scale AI data center set to be built in Ulsan will house world-class GPU computing infrastructure. According to current plans, the facility will launch its first phase of operations in 2027 with 41 MW of capacity and expand to 103 MW (approximately 60,000 GPUs) by 2029. In the long term, the data center is expected to scale up to 1 GW, positioning it not only as Korea’s most powerful computing facility but also as a future hub for AI in Asia.

It is essential that this high-performance computing capacity be made accessible to local manufacturing and logistics companies to actively support their AI transformation. For reference, the National AI Data Center in Gwangju—equipped with 2,184 GPUs (88.5 petaflops)—supported AI development by offering free access to computational resources to over 870 companies and institutions in 2023 alone.

Ulsan should adopt a similar approach by creating an “AI Workload Demonstration Lab” that shares part of the data center’s GPU cluster with the local community and businesses.

Specifically, Ulsan City and the data center’s operators (e.g., SK Broadband, AWS) could establish a cooperative MOU to create an AI testbed for local enterprises. A cloud-based GPU rental service could be developed, allowing local SMEs and startups to access the computing resources needed to develop and train AI models—free of charge or at reduced cost for limited durations.

For example, a “Ulsan AI Voucher” program could be introduced, offering selected companies GPU usage credits. These companies would then access the data center’s high-performance computing (HPC) cluster to run model training and testing using their own data. This initiative could be aligned with national support programs, such as the Ministry of Science and ICT’s AI computing assistance scheme or the existing AI Voucher Program, to ensure broader funding support.

Other local governments have demonstrated similar models. Incheon, for example, provided free access to high-end GPU servers (A100/A6000) for eight months through its AI Convergence Research Center, followed by a paid model at a discounted rate. With its designation as a national AI infrastructure hub, Ulsan is well-positioned to offer even larger-scale and longer-term support.

4.2 Establishing an AI Demonstration Lab and Fostering Talent

In addition to sharing GPU infrastructure, Ulsan should establish an AI Demonstration Lab within or near the data center. Functioning as an open lab, this facility would assist companies in conducting PoC (proof-of-concept) or pilot AI projects.

For instance, if a manufacturing company wants to deploy AI for defect detection, it could bring its data to the lab, receive expert consultation, and develop/test its models on site. Similarly, a logistics company seeking to validate route optimization algorithms could leverage simulation tools provided by the lab. By offering a one-stop support system combining expert guidance, computing power, and test environments, the lab would significantly reduce the barriers for local businesses to adopt AI.

The lab should also operate industry-academia collaboration programs, enabling AI students from local institutions such as UNIST to participate in real business projects. This would help students gain hands-on experience with advanced GPU tools while giving companies opportunities to identify and recruit talent through internships—creating a positive feedback loop for AI workforce development.

Moreover, since Ulsan is also developing specialized zones for emerging industries such as secondary batteries, the AI Lab could be linked to support centers in those zones to generate further synergy. By launching joint R&D projects—such as AI + manufacturing convergence or AI-powered logistics optimization—Ulsan could drive innovation using the resources of the AI data center. Companies that achieve proven results could be rewarded with priority GPU access, investment linkages, or other incentives to enhance participation.

In terms of funding, the city could explore national innovation budgets or leverage cooperation funds from corporate stakeholders operating within the data center. Companies like SK Telecom, which are actively expanding into the global AI infrastructure market, may be encouraged to make social-impact investments that support regional innovation ecosystems.

Ultimately, Ulsan must pursue a proactive and open strategy that transforms the AI data center from a simple server hub into a powerhouse of industrial AI innovation.

5. Sustainable Energy–Based Clustering of Data Centers and Logistics Infrastructure

5.1 Eco-Friendly Data Centers Aligned with RE100 Goals

In the global shift toward ESG-focused development, the energy consumption of data centers has become a critical issue. Ulsan has the opportunity to take the lead by presenting a forward-looking model for eco-friendly AI data centers. As previously mentioned, the Ulsan AI Data Center will leverage electricity from a nearby LNG power plant and utilize cold energy from LNG regasification to improve cooling efficiency—making it one of the most energy-efficient infrastructures in the country.

In the long term, however, the center should pursue a strategy aligned with 100% renewable energy sourcing (RE100). Fortunately, Ulsan is advancing one of the world’s largest floating offshore wind power projects (6.2 GW), with phased operations set to begin in the coming years. Once completed, this wind farm will reduce 9.3 million tons of carbon emissions annually and supply electricity to 5.3 million households in the Yeongnam region. Experts estimate that it could supply enough renewable energy to power every factory in Ulsan’s national industrial complex, making RE100 a realistic and achievable goal.

With this abundant renewable energy potential, Ulsan is ideally positioned to provide clean power to AI infrastructure.

From a policy standpoint, Ulsan should work closely with the data center operators to develop a formal renewable energy procurement roadmap. This can begin with power purchase agreements (PPAs) or the acquisition of renewable energy certificates (RECs), with the share of renewable power gradually increasing over time. Once the offshore wind farms become commercially operational, the data center could source large-scale electricity directly through long-term PPAs—allowing for early achievement of RE100 targets.

This strategy would also benefit domestic and global companies using the Ulsan data center by enhancing their carbon neutrality credentials, thereby positioning the site as a highly attractive ESG-aligned cloud hub. Furthermore, the city could explore circular energy systems, such as reusing waste heat from server rooms to provide heating to nearby industrial facilities or support LNG regasification, while simultaneously utilizing cold energy from the LNG terminal for data center cooling. Optimizing this thermal energy exchange would significantly boost overall efficiency.

5.2 Transitioning to a Carbon-Reduced Logistics Hub

Ulsan, historically rooted in energy-intensive industries like petrochemicals and refining, now has the opportunity to transform itself into a low-carbon logistics cluster centered around the new data center, port, and industrial complex.

Ulsan Port is already restructuring its aging petroleum terminals to handle clean fuels such as LNG, and is planning to evolve into a logistics hub for next-generation fuels like ammonia, hydrogen, and methanol. This shift aligns with global maritime decarbonization trends—and the data center itself will serve as a critical node within this clean energy ecosystem.

Specifically, clean energy produced by future infrastructure at Ulsan Port—such as liquefied hydrogen plants and ammonia terminals—can be used to power the data center through fuel cells. Alternatively, the data center’s stable electricity demand could support the growth of distributed renewable energy projects. Integrating the center’s power infrastructure with onshore power supply (AMP) systems for docked ships or hydrogen-powered port vehicles would further contribute to decarbonizing port operations.

On the logistics side, it is recommended that a carbon tracking module be added to the port–industrial logistics platform mentioned earlier, enabling real-time monitoring and management of emissions generated during transportation. The AI computing capabilities of the Ulsan data center could be used to calculate the carbon footprint of an entire logistics chain—for example, from a factory to a truck, to the port, and onto a vessel.

Such data would form the foundation for effective green logistics policy and help companies prepare for carbon taxes and carbon border adjustment mechanisms. Ulsan could also serve as a testbed for RE100-compliant logistics, allowing companies to pilot the use of electric trucks, hydrogen yard tractors, and eco-friendly cargo cranes powered by renewable energy. AI would optimize their operation, minimizing emissions across the logistics process.

This innovation would elevate Ulsan’s port as a Green Port in international trade and create momentum for attracting sustainability-driven investments.

Most importantly, the development of a carbon-neutral industrial cluster centered on the data center is key to Ulsan’s long-term sustainability. With its high industrial carbon emissions, Ulsan stands to benefit from infrastructure that enables RE100 and low-carbon logistics. These initiatives would be well-aligned with national carbon neutrality funding programs and receive favorable evaluations in global ESG assessments.

As multinational companies increasingly demand decarbonization across their supply chains, Ulsan’s proactive investment in clean energy and infrastructure will make it an attractive destination for firms pursuing green strategies. The vision should be clear: using the Ulsan AI Data Center as a core catalyst, the city must integrate energy, logistics, and industry into a comprehensive green transformation strategy.

6. Conclusion: Ulsan as a Leading Hub for Industrial Safety and Logistics Innovation in the AI Era

Taken together, the four strategies proposed— ① advancing AI-based industrial safety, ② integrating port–industrial logistics through a unified platform, ③ opening GPU infrastructure to support local enterprises, and ④ building a renewable energy–powered data center and logistics cluster— demonstrate that Ulsan is not simply hosting a large-scale AI facility, but is positioned to become a central city driving digital transformation across manufacturing and logistics.

In particular, Ulsan’s approach to integrating AI and data across the safety, logistics, and energy sectors represents a pioneering model rarely seen in other regions of Korea. This makes it a strong candidate for securing government funding and institutional support.

For example, by taking the lead in building AI-based systems for serious accident prediction and real-world safety demonstration zones, Ulsan can secure first-mover policy advantages and national rollout support. Similarly, linking port and inland logistics data positions Ulsan to serve as a pilot city for smart logistics initiatives by the Ministry of Land, Infrastructure and Transport and the Ministry of Oceans and Fisheries. Its push for eco-friendly data centers and logistics hubs could qualify as a national carbon neutrality flagship project, eligible for financial incentives and regulatory exemptions.

The intangible benefits are equally significant. If Ulsan succeeds in these AI-driven innovations, the city’s image as an industrial hub will evolve into that of a safe, smart, and innovative city—attracting both top talent and forward-thinking companies.

Not only large corporations, but also high-potential AI startups will be drawn to Ulsan to test and commercialize their solutions. This, in turn, will generate local jobs, attract young talent, and invigorate the regional economy.

From an ESG perspective, global investors prioritizing sustainable business practices will recognize Ulsan’s industrial complexes as eco-smart zones, enhancing the global competitiveness of the city’s core industries.

In summary, Ulsan must leverage its AI data center as a catalyst for industrial upgrading, pursuing a strategic approach that simultaneously addresses safety, logistics efficiency, and carbon neutrality.

This policy vision will position Ulsan as a national testbed for industrial innovation, allowing it to play a leading role in shaping Korea’s AI transformation agenda.

Once known as the “Capital of Manufacturing,” Ulsan is now poised to become the Capital of AI-Powered Industrial Innovation—bringing new momentum to both the regional and national economy in the coming era of artificial intelligence.

We invite continued support and attention to Ulsan’s bold challenge, as the city works toward becoming a safe, smart, and sustainable industrial hub for the future.

HOSOON CHOI (Hosoon Choi Insight)

Insight from Korea’s Strategic Logistics Frontline

※ Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official position or policy of any affiliated institution or organization.

This content is provided for general informational purposes only and should not be interpreted as legal or commercial advice regarding any specific company, institution, or policy.

Hashtag :

References :

YTN. “울산에 동북아 최대 AI 데이터센터...유치 배경은?” (2025.06.20)ytn.co.krytn.co.kr
동아일보. “SK, 아마존 손잡고 울산에 국내 최대 AI데이터센터 건립” (2025.06.15)donga.comdonga.com
SK텔레콤 뉴스룸. “SK-AWS 울산 AI 데이터센터, AI 고속도로의 거점으로” (2025.06.23)news.sktelecom.comnews.sktelecom.com
뉴시스. “AI, 산업현장 ‘안전지킴이’ 되다…중대재해 예방기술 각광” (2025.03.03)newsis.comnewsis.com
뉴시스. 위 기사 newsis.comnewsis.com (슈퍼브AI, 세이퍼스 등 산업안전 AI 솔루션 사례)
세계일보. “산재사고 5건 중 1건은 울산국가산단…원인과 해법은” (2023.02.15)khan.co.kr
World Port Sustainability Program. “PortWise – Ulsan Port Authority (Intelligent Port Logistics Integrated Platform)” (2025)sustainableworldports.orgsustainableworldports.org
매일경제 Pulse(영문). “South Korea to develop port cargo tracking platform by 2027” (2025.05.19)pulse.mk.co.krpulse.mk.co.kr
울산항만공사 보도자료. “울산항 데이터통합플랫폼 ‘PortWise’ 앱 출시” (2023)ksilbo.co.kr
MBC뉴스. “[단독] 1조 원 들여 GPU 산다더니‥있는 GPU도 절반 놀아” (2025.04.08)imnews.imbc.comimnews.imbc.com
인천시 통합게시판. “[인하대-인공지능융합연구센터] 고성능 GPU 서버 무상지원 안내” (2021.11)incheon.go.kr
울산저널. ““부유식 해상풍력으로 RE100… 구현”” (2025.04.21)m.usjournal.kr
전기신문. “'오일에서 에너지로'…울산신항, 청정에너지 허브로 대전환” (2023.06.05)electimes.com
울산매일신문. “울산 공단에 태양광 발전 확대.. RE100 로드맵” (2023.02.27)kyilbo.com