Hansol Paper said on October 15, 2026 that it had been selected for a government-supported AI-factory programme for the papermaking process, according to Yonhap’s report. Reporting the company’s account, Yonhap described a phased plan to integrate and standardise process, equipment and quality data, then develop and validate models for heat-energy optimisation, equipment anomaly detection and quality analysis. A reported selection and development plan are not measured steam savings.
What the reported programme covers
Yonhap said the consortium comprises seven participants: two demand-side companies including Hansol Paper, three AI specialists and two research organisations. It reported a project running through 2029, worth KRW 12.2 billion in total, including KRW 8.5 billion in government support. These are figures from a report attributing the plan to Hansol. They are neither Hansol’s standalone capital spending nor money already spent. We have not independently examined the government’s underlying award notice, agreement or budget allocation by institution.
The proposed use cases include optimising steam production at an incinerator and reducing steam consumption in the paper-machine drying process. These are development and validation targets. The report does not supply a measured saving rate, heat-input balance, impact on quality or scope of closed-loop operation under production conditions.
Evidence needed for an operating result
Steam demand in papermaking varies with grade, basis weight, machine speed, incoming moisture and downtime. A monthly total alone could confuse a change in production mix with an effect of AI control. Once results are released, useful evidence would include a matched baseline and comparison period, steam use per tonne by comparable grade, moisture and quality constraints, anomaly events, and records of operator intervention. These are our proposed checks, not results announced by the project partners.
Likewise, building an anomaly alarm is different from using it in maintenance decisions and showing fewer breakdowns. False alerts, missed failures, the observation period and actual downtime belong alongside model accuracy. It would also matter whether trials covered conditions where energy savings and product quality pull in different directions.
What packaging buyers should watch
Paper produced by Hansol can enter packaging supply chains, but this announcement establishes no new grade, certification or shorter lead time for any named packaging paper. Buyers can watch for later disclosure of the mill, paper machine, grades and quality controls actually covered by the project. A trial on one machine or grade should not automatically be applied to another.
What is documented now is the company’s selection account as reported by Yonhap. Installation, validation and routine commercial operation are separate future stages; the 2029 horizon should not be mistaken for operating proof.
About the Author
PackingMaster: Editor of PaperPackLog, covering paper-packaging industry developments, products and technical decisions.
References
- Yonhap, report on Hansol Paper’s AI-factory project selection, October 15, 2026 — Reports the company’s account; not independent verification of operating performance.