[AI TECH 2026] Vision AI, Sound AI, and Autonomous Manufacturing: Key Conditions for AX Success Across Industries

Enterprise AI adoption is rapidly expanding across manufacturing, logistics, and construction, but cases that go beyond PoC and deliver tangible results remain limited.

The reason is that companies often struggle to collect data properly in the field, or AI models that work in laboratory settings fail to reproduce the same accuracy on real production lines. In other words, the gap between “adopting AI” and “achieving results with AI” is widening.

To address these practical challenges and discuss the concrete conditions for successful AI transformation in industrial settings, the AI Convergence Business Development Conference (AI TECH 2026) will be held on May 6 at Hall E, 3rd floor, COEX in Seoul.

This year’s theme, “Make AI Work for Your Business,” highlights how AI can be made to work in real-world operations.

The C Track of AI TECH 2026 is themed “Successful AI Convergence Cases by Industry” and will focus on how AI is creating real change in major industrial sectors such as manufacturing, logistics, and construction.

The first presentation will be delivered by Lee Min-hye, Head of the AI Division at VB Company, under the title “AI Agent Adoption Cases and Reality: Conditions for Becoming a Corporate Competitiveness Driver.” Drawing on the experience of a first-generation AI and big data company with the largest number of AI PoC cases, she will explain how AI agent technology is evolving beyond LLM chatbots and RAG-based AI into task-complete “Deep Agents,” and will introduce actual deployment cases in public institutions and financial organizations. She is expected to discuss why corporate AX efforts fail, citing issues such as the confusion between using AI and transforming with AI, AI use that remains tacit knowledge, and the lack of systemization, while also presenting practical action plans to spread AI capabilities across the organization.

Next, Lee Jae-min, Head of Business at Superb AI, will present on “Real AX Trends and Application Cases Accelerated by Vision AI in Manufacturing, Logistics, and Construction.” Centered on three pillars—standard operating procedure analysis, safety monitoring via video analytics, and defect detection through an MLOps platform—he will explain how machine vision AI is replacing manual worker logs and establishing real-time risk detection systems in manufacturing, with real-world examples. He will also share methods for continuously improving AI performance, including an MLOps feedback loop in which operator feedback is automatically converted into training data.

Following that, Choi Hyeong-jun, Team Lead of AI Execution Planning at M. Cube Solution, will present “Evolution of Manufacturing Systems for the Era of Autonomous Manufacturing: Practical Strategies for Defect Detection and Predictive Maintenance.” As the shift from DX to AX accelerates, he will examine the structural limitations of existing MES and FDC systems and propose a closed-loop digital twin implementation strategy combining agentic AI and physical AI. He will also cover specific field cases, such as the integration of LLM and digital twin in a steel manufacturer and the real-world results of an LLM multi-agent system in a secondary battery EMS environment.

The final presentation will be delivered by Suji Lee, CEO of Deeply, under the title “Physical AI Implemented with Sound AI: The Final Piece of the Factory Automation Puzzle.” Starting from the recognition that while vision-based quality inspection has largely been automated by machine vision AI, auditory inspection for motors, speakers, and fastening sounds still depends entirely on human labor, she will present cases where the deep learning-based anomaly sound detection solution Listen AI Industrial achieved more than 60% annual labor cost savings per line and 99% inspection accuracy. She is expected to emphasize that machine hearing, following machine vision, is emerging as the next AI transformation area in manufacturing.

An AI Tech representative said, “Corporate AI transformation is not simply about adopting a specific technology; the key is which on-site problem it will solve and how the results will be sustained.” The representative added that this C Track will be a concentrated venue of experience from experts who have directly implemented AI in industrial settings and created measurable outcomes.

Meanwhile, AI TECH 2026 will be held on May 6 from 10:00 a.m. to 4:50 p.m. at Hall E 1–4 on the 3rd floor of COEX in Seoul. The morning session will feature an opening keynote with KAIST, Microsoft Korea, Code Fantasia, and M Cloud Bridge, while the afternoon will be divided into four tracks: Agentic AI & Business Innovation (Track A), AI Development to Operations (Track B), Successful AI Convergence Cases by Industry (Track C), and Physical AI Implementation Strategies (Track D). Attendees may move freely between sessions and meet experts on topics of interest.

AI Tech is a paid event, and early-bird registration as well as a 30% discount for groups of three or more are currently available. Pre-registration can be completed on the Dubiz website.