How Top Teams Accelerated Marine Data Compression in the Ship‑Sea Innovation Contest

Ship‑Sea Data Intelligent Application Innovation Competition Overview

The first "Ship‑Sea Data Intelligent Application Innovation Competition" was organized by Taihu Laboratory, the Wuxi Municipal Talent Work Leadership Group Office, the Wuxi Science and Technology Bureau, and Huawei to promote the integration of modern information technology, next‑generation AI, and ship technology.

Competition Focus

In the second track of the preliminary round, titled "Unstructured Data Compression and Processing," participants tackled the problem of efficiently storing and transmitting the massive amount of unstructured data generated by marine scientific experiments, which accounts for over 80% of total data and poses high storage‑cost challenges.

Top Performers

First place was claimed by Wang Xingyue from Shanghai Maritime University, second place by Mo Haige from the University of Science and Technology of China, and third place by Xiao Jinzhou from Tsinghua University.

Winning Solutions

Wang Xingyue described a two‑step approach: first preprocess the data to group regular patterns for compression, then treat irregular parts as checksum data and apply a verification‑oriented compression algorithm, ultimately using a standard compression method to achieve higher ratios.

Mo Haige emphasized classifying and encoding the dataset before applying mature general‑purpose compression algorithms, noting that the difficulty lies in the initial categorization of unstructured data.

Xiao Jinzhou experimented with various encoding schemes (PLAIN, TS_2DIFF, RLE, SPRINTZ, GORILLA, RLBE, RAKE) combined with compression algorithms (LZ4, SNAPPY, GZIP). The team selected RLE with bit‑packing plus GZIP as the base, then refined it for binary and BCH‑code columns, achieving a substantial compression improvement.

Challenges and Support

All three teams encountered obstacles related to data irregularities and read/write correctness. They resolved these issues through literature research, peer discussion in the competition forum, and technical guidance from the organizers.

Impact and Future Outlook

The competition highlighted the strategic importance of marine data digitization, attracted high‑caliber talent, and demonstrated how industry‑academic collaboration can foster innovative ecosystems that accelerate digital transformation in maritime and related industrial sectors.

Participants expressed that the contest serves as a valuable platform for talent recruitment, professional networking, and future collaborations, while emphasizing the need for more challenges focused on large‑scale industrial data scenarios.