알림 목록

Analysis Area: Daegu Metropolitan City

Core Areas: Seongseo Industrial Complex, Dalseong Industrial Complex, Suseong Alpha City, Daegu National Industrial Complex, and areas with a high concentration of local manufacturing companies

Agenda: Are the M.AX Declaration and the Industry AX Innovation Hub leading to changes in productivity, defect rates, delivery times, and costs for local manufacturing companies?

Golden Time Type: Execution Gap + Structural Transformation

Reference Date: August 28, 2026

Version: Regional AX Golden Time Intelligence v3.2
 

Daegu City has placed the AI ​​transition of the manufacturing sector at the forefront of its industrial policy by 2026 and is promoting the Suseong Alpha City Industrial AX Innovation Hub and the Regional Hub AX Innovation Technology Development Project. However,  Daegu City’s assessment that over 90% of local companies are in the early stages of AI adoption, coupled with large-scale hub investments scheduled until 2029, serves as evidence that a significant gap exists between the policy's starting point and the level of readiness on the ground.  While hub construction and R&D investment will proceed over the next two to three years, delays in digitizing corporate processes will result in a disconnect between the timelines of facilities and the actual site. Although  Daegu M.AX has moved from the declaration stage to the investment stage, evidence to assess this as a structural change in the local manufacturing sector remains limited.

The Industrial AX Innovation Hub combines research and development with demonstration, focusing on robotics and mobility, brain disease healthcare, and intelligent semiconductors. In contrast, Daegu's manufacturing sector has a multi-layered structure where production and employment are supported by numerous small and medium-sized enterprises (  SMEs) in sectors such as automotive parts, machinery, metal, textiles, and eyewear. If a diffusion pathway connecting the strategic industry-centered AX with the process transitions of existing manufacturing companies is not identified, the productivity gap between leading companies and the rest of the manufacturing industry will widen by 2028–2029.  Currently, the primary criterion for evaluating Daegu M.AX is not the scale of the hub, but rather the depth of diffusion extending down to traditional manufacturing companies .

Daegu City, in collaboration with DGIST, is constructing  the Industrial AX Innovation Hub in Suseong Alpha City with a total project cost of 47.7 billion won , investing 39.6 billion won in state funds and 8.1 billion won in city funds . With a schedule set for site acquisition in 2026, groundbreaking in 2027, and completion in 2029, Daegu's AX policy has shifted from individual support projects to the establishment of a permanent research and demonstration base. However, criteria for disclosing joint AI models to be supplied to local companies, manufacturing data usage rights, and individual company conversion outcomes prior to the hub's completion remain unconfirmed. If on-site application remains on hold until the facility is completed in 2029, the current 90% of early-stage companies will remain as demand-side enterprises even three years from now.  While the establishment of physical infrastructure has been confirmed, manufacturing conversion outcomes have not been determined in advance.

The government's Regional Hub AX Innovation Technology Development Project envisions fostering Daegu as a hub for AI, robotics, and semiconductor convergence R&D and industrial demonstration, with a total project budget of approximately 551 billion won over five years starting in 2026. By combining a physical hub worth 47.7 billion won with technology development worth 551 billion won, Daegu's M.AX has expanded from a standalone local project into a key pillar of the National AX Innovation Belt. However, since the 551 billion won is presented as the total scale of the regional hub project rather than a single project budget fixedly allocated to Daegu alone, the amount attributable to the region and the results attributable to companies must be interpreted separately. If the total project budget is repeatedly cited as regional achievement over the next two to three years, the actual amount invested in Daegu companies and the results may be overestimated.  While the financial scale of Daegu M.AX has increased, the actual impact attributable to the local manufacturing sector remains undetermined.

Daegu is pursuing the autonomous manufacturing project for implantable medical devices with a total budget of 11.8 billion won from 2024 to 2028 , and was selected in 2025 for the AI ​​Accelerator Ultra-High Multilayer PCB Autonomous Manufacturing Technology Development Project, with an investment of 10.6 billion won until 2029. As Daesung Hitech and Isu Petasis demonstrated medical device and PCB processes, respectively, M.AX has moved from the conceptual stage to the actual manufacturing line.  However, both projects are centered on mid-sized and leading companies with high technological capabilities and investment capacity, and track records of transfer to numerous small-scale manufacturing enterprises have not yet been confirmed. If the performance of the leading factories does not lead to the expansion of standard models and supply chains by 2028, M.AX will remain confined to being an advanced project for a select few companies. While  actual factory transformation has begun, it is premature to conclude that it has spread to transform the entire regional manufacturing sector.

The Industrial AX Innovation Hub aims to begin construction in 2027 and complete it in 2029, while the Regional Hub AX Innovation Technology Development Project will run for five years starting in 2026.  A supply-first structure has been established where buildings, equipment, and research organizations are fixed first, followed by the demand for data preparation and adoption from manufacturing companies.  If more than 90% of local companies are in the early stages of AI adoption, significant time is required for the stages of collecting, standardizing, and linking process data; however, there is no evidence breaking down this level of preparation by industry and company size. If the speed of technology development outpaces the speed of field data preparation by 2028, companies capable of conducting demonstrations will be limited to leading firms.  Daegu M.AX operates under an initial structure where the supply of research and infrastructure precedes the demand from manufacturing sites.

The AX Industry Division at Daegu Technopark is responsible for integrating AI technology into aging manufacturing processes while carrying out numerous projects, including the dissemination and expansion of smart factories. Although existing smart factory initiatives and the new M.AX initiatives lie on a continuum,  the percentage of companies that have adopted automation and MES and have advanced to AI-based prediction, judgment, and autonomous control stages is not confirmed in publicly available data.  If smart factory installation records are considered indicative of companies preparing for M.AX, data quality and process connectivity levels are overestimated. Over the next two to three years, as existing dissemination records and new AX achievements are mixed, it will become difficult to separate the actual conversion rate.  While Daegu possesses a digital manufacturing infrastructure, the scope verified as a foundation for M.AX conversion remains unclear.

The autonomous manufacturing project for ultra-high-layer PCBs for AI accelerators is a national AI factory pilot project that applies AI-based cognition, judgment, and control to manufacturing processes. While a structural transformation has begun in Daegu's manufacturing sector, integrating the AI ​​semiconductor supply chain with autonomous manufacturing within a single factory,  the contract, standard, and supplier structure for transferring Isu Petasis's process performance to local PCB, machinery, equipment, and material companies has not yet been confirmed. If only the productivity of the lead company increases by 2029, the AX gap within the regional supply chain will widen.  What has been confirmed so far is the transformation of the pilot factory, not a transformation of the regional manufacturing ecosystem.

The autonomous manufacturing project for implantable medical devices, centered around Daesung Hitech, connects precision machining and medical device manufacturing with AI processes.  While a pathway has been opened for machinery and parts manufacturing capabilities to shift to high-value-added processes in the medical industry , the number of diffusion cases where local machinery companies applied the same model and productivity after adoption have not been disclosed. If the technology remains confined to a single factory and participating institution until the project concludes in 2028, the effect of industry transformation on the local manufacturing sector will be limited.  Daegu M.AX is currently in the demonstration phase of inter-industry integration and has not yet reached the generalization stage.

M.AX cannot advance to prediction, optimization, and autonomous control without data linking equipment status, quality, working conditions, energy, and production plans. The fact that Daegu City diagnosed over 90% of local companies as being in the early stages of AI adoption is indirect evidence that the level of connectivity between data and systems is also likely in its early stages.  However, data collection rates by industry, MES/ERP connectivity rates, equipment communication capability ratios, and the retention period for trainable data have not been disclosed. Even if model development is completed first in 2027–2028, a lack of field data will increase customized implementation costs and slow the speed of diffusion.  The biggest leading gap for Daegu M.AX is the level of manufacturing data readiness, rather than the AI ​​model itself.

Daegu's manufacturing sector consists of industries with diverse processes and equipment generations, such as automotive parts, machinery, metals, textiles, and eyewear. Despite this structure, which makes it difficult to apply standardized platforms or AI models, evidence regarding established industry-specific data standards and the scope of collaborative learning is limited. If data structures for individual projects remain fixed for two to three years, interoperability and reusability will be diminished after the projects conclude. If data is fragmented into company-specific and project-specific assets, 551 billion won in R&D investment will not be accumulated as a shared regional asset.  Currently, there is a high risk that Daegu's manufacturing data will be formed as project-specific resources rather than based on an industry-wide foundation.

The pilot projects are being implemented as multi-year consortia worth 10.6 billion KRW and 11.8 billion KRW, with participating companies receiving support from both research institutions and specialized AI firms. While this structure is suitable for technology development, it differs from the actual adoption conditions of small manufacturing companies that lack independent data personnel and investment capacity.  Evidence regarding payback periods and total costs of ownership—including corporate contributions, maintenance costs, cloud and computing expenses, and equipment modification costs—by industry sector has not been verified. If market prices are not established outside of subsidy programs by 2028, voluntary diffusion will stagnate after support ends.  While Daegu M.AX is demonstrating technological feasibility, it has not yet reached the stage of proving economic viability for small and medium-sized enterprises.

A majority of manufacturing companies in Daegu are located in contract manufacturing and parts supply chains rather than as finished product brands. In a structure where ordering companies determine unit prices and volumes, there is no guarantee that productivity improvements will translate into profits for the adopting companies.  Public data on regional cases linking improved productivity and defect rates with increased operating profits and orders is limited.  If AX investment costs are borne by SMEs over the next two to three years and cost-saving benefits are passed on to lower unit prices, the incentive for adoption will weaken.  The diffusion gap for M.AX in Daegu lies in the structure of investment return attribution rather than the technology itself.

Suseong Alpha City is Daegu's technology hub where software and ABB companies are concentrated with digital infrastructure, and the Industrial AX Innovation Hub is also being established in the same area. While the spatial concentration of AI suppliers and research institutions is strengthening, major manufacturing sites remain dispersed across various industrial hubs, such as the Seongseo, Dalseong, and National Industrial Complexes.  The number of projects where technology companies in Suseong Alpha City stationed themselves at manufacturing sites to resolve process issues, as well as the contract renewal rates, remain unconfirmed.  Over the next two to three years, if suppliers focus on technology development tasks while manufacturing companies cannot afford on-site maintenance costs, the distance between the two parties will widen due to business structures rather than geography.  Although Daegu is establishing AX supply bases, a field-penetrating delivery system has not yet been verified.

Daegu Technopark’s AX support and the Industrial AX Innovation Hub, DGIST’s R&D, and the National AI Factory project are all expanding simultaneously. While supply capacity grows as the number of supporting institutions and projects increases, there is a possibility that channels for diagnosis, data maintenance, demonstration, financing, and commercialization may become fragmented for companies.  The distribution of participation—where the same company participates in multiple projects or, conversely, where early-stage companies fail to enter any—has not been disclosed. If performance indicators for each institution are fixed based on the number of participating companies by 2028, both overlapping support and blind spots will occur simultaneously.  Although the supply system of Daegu M.AX is expanding, the integration pathways at the corporate level remain unconfirmed.

The Ministry of Trade, Industry and Energy's AI Factory project is structured to enhance productivity and quality through AI-based perception, judgment, and control, and the results of autonomous process operation and predictive maintenance are being verified in national M.AX cases. Although Daegu has also secured two pilot projects, a regional performance table disclosing production volume, defect rates, Overall Equipment Efficiency, delivery times, energy costs, and working hours before and after implementation has not yet been confirmed.  If project selection and investment amounts substitute for productivity evidence, the boundary between implementation and results disappears. If baselines for 2028–2029 are not disclosed, comparability becomes low even if improvement rates are presented.  In the case of Daegu M.AX, investment results have already become visible, while productivity performance remains awaiting verification.

Daegu’s industrial statistics provide data on manufacturers, employees, production value, value added, and the industrial production index. However, since changes in macroeconomic manufacturing indicators are influenced by the economy, exports, raw material prices, and investments by large corporations, the effects of the M.AX project cannot be isolated. Panels comparing beneficiary and non-beneficiary firms regarding productivity, employment, and sales are not available in publicly available data.  Even if regional manufacturing production increases over the next two to three years, if the contribution of AX cannot be separated, the policy may claim to have achieved results, but verification will be impossible.  Currently, while manufacturing statistics exist in Daegu, data for measuring the effects of M.AX is absent.

The 47.7 billion won Industrial AX Innovation Hub, regional technology development worth approximately 551 billion won, the 10.6 billion won PCB AI Factory, and the 11.8 billion won autonomous medical device manufacturing are confirmed simultaneously. While  investment, organizational, and demonstration projects have entered the implementation phase, the self-assessment by over 90% of local companies that they are in the early stages of AI adoption reveals that on-site readiness is failing to keep pace with the speed of policy implementation. If a data foundation and internal workforce are not established by 2028, the scope of benefits from large-scale projects will be limited to leading companies. Current readiness levels are assessed as High for public investment, High-Middle for leading companies, and Low for general manufacturing companies .

The autonomous manufacturing projects targeting the actual manufacturing lines of Daesung Hitech and Isu Petasis serve as evidence that factory application has begun . On the other hand, the number of cases where these project models were reused by local partners or similar SMEs, as well as the track record of paid adoption, have not been confirmed. If secondary and tertiary diffusion does not occur within 2 to 3 years, a structure centered on leading companies will become entrenched. The level of corporate diffusion for Daegu M.AX is assessed as having initiated demonstrations and that horizontal diffusion has not been confirmed .

The policy targets have been specified as robots and mobility, brain disease healthcare, and intelligent semiconductors, and demonstrations of medical device and PCB processes are also underway. However, the achievement of AX spreading to Daegu's existing manufacturing base, such as textiles, eyewear, metal processing, and general machine parts, is limited. If the connection between strategic industries and traditional industries remains weak by 2028–2029, existing industries will fall behind in both support priority and competition for technical personnel. The level of industry expansion is assessed as being concentrated on strategic industries and in the early stages of traditional industries .

The AI ​​supply capabilities of DGIST and Suseong Alpha City, along with Daegu Technopark's smart factory and AX support organizations, serve as evidence of the existence of an external foundation of expertise. Conversely, the level of data managers, process-AI convergence personnel, and long-term training data retention within manufacturing companies has not been disclosed. If reliance on external implementing agencies continues for two to three years, capabilities for model operation, retraining, and error response will not accumulate as internal corporate expertise. The level of data and personnel readiness is assessed as above-average for supply agencies and below for internal companies .

Construction of the Industrial AX Innovation Hub is scheduled to begin in 2027 and be completed in 2029, while the two AI Factory projects will also proceed from 2028 to 2029. Although the period from 2026 to 2028 is a phase where manufacturing data, common models, and corporate diffusion pathways are determined prior to facility completion, current achievements are focused on project selection and infrastructure development. If paid adoption and repeated diffusion by general manufacturing companies do not occur during this period, the Hub will launch in 2029 as a large-scale public supply facility supporting startups. The golden time for the practical transition of Daegu M.AX is estimated to be the next 24 to 30 months .

The site, design, and construction of the 47.7 billion won hub, along with the five-year technology development projects, fix assets, organizations, and implementing agencies for an extended period.  Once corporate data becomes fragmented by project and a support structure centered on leading companies is established, the cost of modifying the allocation structure to center on general manufacturing companies increases. Furthermore, after the facility is completed in 2029, the likelihood of securing an operational rate taking precedence over performance evaluation also rises. Irreversibility is assessed as high risk for physical infrastructure, medium-to-high risk for technology projects, and high risk for the gap with general companies .

Daegu M.AX did not stop at mere declarations.  It moved beyond securing national projects, finalizing the budget for the Industrial AX Innovation Hub, and demonstrating an AI factory on an actual manufacturing line.

However, the scope of confirmed change is limited to public infrastructure and a few leading firms . Diffusion evidence has not yet been formed to overturn the starting condition that more than 90% of local manufacturing firms are in the early stages.

The final assessment is “M.AX Infrastructure Acceleration + Manufacturing Diffusion Gap .  ” Although Daegu has actually launched the M.AX project, it is not yet at the stage where it can be judged to be actually transforming the entire regional manufacturing sector.

Overall Score: 52 / 100

Golden Time Status:
ORANGE

Time Window:
24~30 months

Failure Mode:
Large AX hubs and a few leading plants remain, but the majority of manufacturing companies remain in the initial stage.

Irreversibility:
Infrastructure and project structure medium-risk, manufacturing firm gap high-risk

Evidence Sources

• National Artificial Intelligence Strategy Committee—On-site Inspection of Regional Hub AX Innovation Technology Development Project
• Ministry of Trade, Industry and Energy—On-site Achievements of the AI ​​Transformation in Manufacturing
• Ministry of Trade, Industry and Energy · Financial Services Commission — M.AX Frontier Project
• Daegu Industrial Economy Trends—Regional Manufacturing Statistics DB
• Daegu Industrial Economy Trends—Mobility Industry Manufacturing Indicators
• Daegu Technopark AX Industry Headquarters
• Daegu Industrial AX Innovation Hub Establishment Plan
• Daegu AI Factory Leading Project—10.6 billion KRW Autonomous PCB Manufacturing
• Daegu Policy Research Institute—AX Capital Daegu Model and 8 Major Strategies

Structural Insight — The success or failure of Daegu M.AX depends on 'replication costs' rather than the number of companies adopting AI

AI in leading factories is implemented tailored to the respective company's equipment, processes, data, and operating conditions. Even if productivity is verified at one factory, if the next company restarts data collection and model development, diffusion becomes a new R&D project every time.  Under this structure, even if the number of success stories increases, the speed of transformation in the regional manufacturing sector grows only linearly.

Daegu is a manufacturing city characterized by significant process heterogeneity, such as in machinery, automotive parts, textiles, eyewear, and medical devices. While it is difficult to integrate these diverse processes into a single AI model, there is no reason for diagnostic methods, data specifications, equipment connectivity, performance measurement, and security contracts to vary every time.  Unless these common elements remain as standard assets, the 47.7 billion won hub and 551 billion won technology development will result in numerous isolated demonstrations.

Therefore, the structural turning point of Daegu M.AX lies not in the performance of the first factory, but in how much the construction costs and time for the second and third factories decrease . If replication costs do not fall rapidly, M.AX will fail in the transition of regional manufacturing even if it is technically successful. Daegu's actual speed of transition is determined not by the total investment, but by the rate of decline in the marginal cost of replacing one more factory .