Sound AI That Catches Defects Is Drawing Startup Attention in Manufacturing Sites

Domestic deep-tech startups are accelerating the AI transformation of traditional manufacturing by adding “eyes,” “ears,” and “brains” to industrial equipment. The use of unstructured data in real time for perception, analysis, and control is changing both process efficiency and safety management across production sites.

According to the venture capital industry on the 7th, SuperB AI, a vision AI startup, is actively expanding into the Japanese market. Its technology is already known to have been adopted in manufacturing sites by companies including Toyota, Denso Ten, and Nippon Steel.

SuperB AI built its technology around ZERO, a vision foundation model pretrained on massive image data, enabling factories to instantly recognize defective products without separate data labeling. The company has also expanded into more advanced areas such as automated data segmentation and quality control model development, improving process efficiency.

It is also expanding into safety management. Using video monitoring solutions based on existing CCTV infrastructure, it can detect missing helmets or safety vests, unauthorized entry into hazardous zones, workers collapsing, and fires or smoke in real time.

Based on this, SuperB AI plans to expand its customer base to Japanese national research institutes and major electronics manufacturers, while strengthening its position in the global manufacturing automation market. In the U.S., it is broadening the scope of its solutions through partnerships with Hanwha Vision America and Advantech.

Deeply is advancing quality inspection methods in manufacturing through acoustic AI. Its core innovation is converting inspections that previously depended on workers’ senses or anechoic chamber environments into data-driven analysis for actuator sounds, motor sounds, and wire connector fastening sounds.

Deeply’s acoustic AI solution, ListenAI, is linked with robotic arms and manufacturing execution systems (MES) on mass-production lines and can determine quality in less than one second per unit. Recently, it has also been introduced in the fastening process of “safety boxes” designed to prevent vehicle battery fires, expanding into industrial safety by detecting missing or incomplete fastening.

A representative domestic case is Hyoseong Electric. In February, Deeply supplied Hyoseong Electric with its industrial acoustic analysis solution, Listen AI Industrial. The company plans to extend the same solution not only to the existing plant but also to a new plant of about 9,000 pyeong (29,752 square meters). In the future, it aims to expand beyond auto parts to robot, mobility, and electric vehicle actuators.

Deeply is also speeding up global expansion. It has supplied component quality and fastening sound inspection solutions to a global automaker’s plant in Mexico. The company has also provided acoustic AI technology to the Thailand Automotive Institute and plans to further expand overseas sales channels by participating in Automate Show, North America’s largest automotive industry exhibition, in Chicago.

Technology that integrates and analyzes data across the factory to make decisions autonomously and optimize operations is also being introduced in manufacturing sites, improving operational efficiency. A representative example is AccelVeo, an intelligent manufacturing solution introduced by Bespin Global in the North American market.

AccelVeo, based on agentic AI, has been deployed at U.S. manufacturing company W. It integrates and analyzes production, maintenance, and supply chain data in real time, detects equipment anomalies in advance, and minimizes downtime. It also improves on-site operating efficiency by automatically adjusting task priorities through AI agents.

Bespin Global aims to move beyond digital transformation in manufacturing and toward the realization of “autonomous factories” through this solution. The company also plans to expand deployments for global manufacturers, especially in North America.

Market interest in technologies that accelerate AI transformation in manufacturing is also growing. A venture capital industry official said industrial AI technologies are attracting attention because their biggest strength is their broad applicability across Korea’s key industries, including semiconductors, automobiles, defense, and shipbuilding. The official added that they are being highly valued not only for technical capability but also for their scalability potential, making them one of the most closely watched sectors in the industry.