Technology Clusters in Paper Industry

Background and results

We have collected and retrieved the active patent portfolios of 13 major paper manufacturers. We analyzed the full patent portfolio of UPM-Kymmene, Stora Enso, Westrock, Huhtamaki, Weyerhäuser, Smurfit Kappa, Metsä and April (Asia Pacific Resources) plus Valmet, Andritz, Kemira and Georgia Pacific (in Koch Industries) (of the latter we have analyzed only patents in patent class IPC D21.) In a next step we applied a modern clustering algorithm using generative AI to cluster and sort the thematically close patents into 61 cluster. The generative AI constructed a description for each cluster, and concluded from the many underlying terms in each cluster a common concept. These are then given as label, which is used as a general descriptor. No further information apart from the original patent texts was used.

The field of paper manufacturers comprises a multitude of aspects, from wood resource management, to pulp handling, paper making, handling and finishing, but also production of end user products, such as food packaging, medical or hygiene products. When analyzing the areas of highest trend character and overproportioned patent publications in recent years, we can clearly find a common concept in the fiber handling, moulding, forming, but also in waste handling and chemical and physical processing. Overall we can see that the paper industry is in an ongoing transformation from a pure producer of flat paper into an industry of three-dimensional and complex articles for a multitude of products and applications, by using natural fiber sources, caring about waste use or reduction and incorporating new processes.

Methodology

The chart shows all 61 defined clusters as bubbles. The bubble size reflects the number of active patents in the cluster. The chart should be read as a gauge where bubbles on the left have very low growth rates (even negative) and bubbles on the right have high growth rates. The colors highlight this: red bubbles have negative growth, yellow have growth rates less than 20% and green above 20%. Growth is determined by the sum of new patent publications in the years 2022-2024 compared to the sum of the patent publications in the years 2019-2021. The clusters are also sorted by the number of focus companies active in each cluster. Clusters in the inner circle have fewer active companies, whereas clusters in the outer circle have all focus companies active.

The top left drop down menu allows to filter the bubbles by company. If a specific company is chosen, only those clusters are shown where this company is active. The color code also changes. Dark blue clusters indicate that the company has relatively fes patents in this cluster (calculated as share of total company patents in the paper industry set. Light blue bubbles indicate a high activity or specialization in this cluster.

Hovering over a cluster or clicking reveals further information on the cluster such as total active patents in the cluster, the change of publications in recent years which determines the growth rate and the number of recent patents (2019-2024) of each active company in the cluster. The GenAI generated label and the cluster description can also be seen.