MapLibre Tile: A Next Generation Vector Tile Format

TL;DR

The paper introduces MapLibre Tile (MLT), a new vector tile format that significantly improves compression and decoding speed over the widely used MVT, supporting next-generation map rendering capabilities.

Contribution

MLT is a novel vector tile specification designed to overcome MVT limitations, offering better compression, faster decoding, and enhanced functionalities for advanced map rendering.

Findings

Up to three times better compression ratios than MVT.

Decoding speeds up to three times faster.

Supports next-generation GPU-accelerated map rendering.

Abstract

The Mapbox Vector Tile (MVT) format is widely considered the leading open standard for large-scale map visualization, as evidenced by its widespread adoption by major technology companies such as AWS, Meta, and Microsoft for their products and services. However, MVT was developed nearly a decade ago and, consequently, does not fully align with the capabilities of new geospatial data sources that are characterized by rapidly increasing data volumes due to advancements in geospatial sensors and automated detection through artificial intelligence. In this paper, we introduce the MapLibre Tile (MLT) format, a novel vector tile specification designed from the ground up to address the limitations of MVT. Our experiments, simulating user sessions on widely used basemap datasets, demonstrate that MLT achieves up to three times better compression ratios compared to MVT on encoded tilesets, with over six times better on certain large tiles. Additionally, MLT offers decoding speeds that are up to three times faster and significantly enhances processing performance. MLT also introduces new functionalities and is specifically designed to lay the foundation for the next generation of map renderers, which we expect to entirely offload processing to the GPU, thereby overcoming the stagnation of Moore`s law.