- Korea Food Research Institute to build autonomous noodle production line

- Sensors read dough conditions; AI decides, robots act

- Productivity up 30%, product loss down 20%

A hand-pulled noodle being stretched. [AI-generated image]
A hand-pulled noodle being stretched. [AI-generated image]

The era of AI replacing the skilled hand of a ramyun or noodle maker is on the horizon. Under the envisioned system, sensors continuously read the moisture and elasticity of dough in real time, AI determines the optimal production conditions, and robots pull the noodles themselves. The goal is to build a fully autonomous noodle production line — one requiring minimal human intervention — by 2030.

The Korea Food Research Institute announced Thursday that it has launched a project to develop autonomous production technology covering the entire noodle manufacturing process by 2030. The project, titled "Development of Intelligent Production Equipment for an Autonomous Production Process for Ultra-Automation of Noodle Manufacturing," is funded under the Ministry of Science and ICT's strategic research program for government-funded research institutes.

South Korea's noodle market — spanning ramyun and other varieties — is growing rapidly. Production value reached about 4 trillion won ($2.95 billion) in 2025, with output hitting 1.4 million tons. Exports surged 22.1 percent, rising from $1.12 billion in 2023 to $1.36 billion in 2024.

Yet outside the automated ramyun lines of large conglomerates, most noodle manufacturing still relies heavily on the experience of skilled workers.

Noodle production involves multiple stages — mixing, kneading, resting, rolling, cutting and shaping — each of which affects moisture, elasticity and texture differently. Even minor shifts in workshop temperature or humidity can alter the dough, requiring fine adjustments to rolling thickness and resting conditions. Applying only preset values, as conventional automated equipment does, makes it difficult to maintain consistent quality.

A conceptual diagram of the autonomous noodle manufacturing process. [Provided by Korea Food Research Institute]
A conceptual diagram of the autonomous noodle manufacturing process. [Provided by Korea Food Research Institute]

The Korea Food Research Institute is now moving to replace those skilled-worker judgments with AI. The core of the system is an autonomous production chain linking sensors, AI, robots and a digital twin.

Multi-function sensors measure dough and noodle quality in real time, feeding data to an AI that predicts quality outcomes and determines optimal operating conditions. Robots handle high-precision tasks such as gripping and shaping noodles. A digital twin — a virtual replica of the actual production floor — monitors and controls the entire process in real time.

The institute is leading a consortium of nine organizations from industry, academia and research. It will oversee noodle quality analysis and process design while integrating the sensors, AI, robots and digital twin into a single production line.

Real-time quality management is expected to cut product loss by more than 20 percent compared with current levels, while autonomous operation should boost productivity by more than 30 percent. The institute plans to raise the automation rate across the entire noodle production process to above 50 percent.

The project will also pursue domestic development of intelligent controllers and robotic equipment that South Korea currently imports in large quantities. Once an autonomous production model is established for noodles, the institute plans to extend it to other food manufacturing sectors such as confectionery and baked goods.

"Noodle manufacturing is a process that relies heavily on skilled workers because the state of raw materials and dough changes constantly," said Im Jeong-ho, a researcher at the institute. "We will complete an autonomous production model where people, AI and robots work together, easing the labor shortage on food manufacturing floors and strengthening the industry's competitiveness."


nbgkoo@heraldcorp.com