The RoboAtlas: Mapping the Global Robotics Landscape
Jiacheng Zhang, Shuo Sun, Vicky Charisi, Xinru Wang, chen xinyue, Zhexuan Ma, Alok Prakash, Thomas Malone
AI summary
Problem
Existing robotics data lacks consistent, model-level details linking robot systems to capabilities, industries, and geography, limiting empirical research on technology diffusion and ecosystem structure.
Approach
The authors developed a three-stage, LLM-assisted pipeline that searches the open web, iteratively verifies company and model data, and extracts structured attributes like robot type, release year, target industries, and tasks.
Key results
- Mapped 8,229 robot models across 1,062 companies in 50 countries
- Identified strong geographic concentration in the US, China, and Japan
- Documented rapid market growth accelerating after 2017
- Cataloged 24,585 unique tasks showing robotics diffusion into healthcare, logistics, and education
Why it matters
Provides a scalable, open-web methodology for tracking global robotics trends and offers empirical insights for researchers, policymakers, and industry stakeholders.
Abstract
Structured, model-level information on the world’s robot systems remains scarce: existing reports often provide aggregated market statistics, while industry directories typically stop at company- or application-level information. In this work, we present an LLM-assisted, web-grounded analysis pipeline for studying the global robotics landscape at the robot-model level. The method combines company discovery, iterative verification, and model-level extraction of robot type, target industries, release year, and task descriptions from open-web evidence. Applying this pipeline, we study 8,229 robot models associated with 1,062 companies across 50 countries and 6 continents. Our findings reveal strong geographic concentration in the United States, China, and Japan, rapid growth after 2017, and substan- tial diffusion of robotics beyond manufacturing into logistics, healthcare, education, and household settings. Interestingly, this analysis revealed 24,585 tasks. Our work illustrates both the promise and certain limitations of LLM-assisted web analysis for large-scale robotics landscape mapping.