Factory Workers and AI Risk in 2026: Robotics Data

Factory assembly sub-roles ranked by robotics risk

ILO ILOSTAT (CC BY 4.0) places assemblers in ISCO-08 group 82. The group includes workers who assemble components in manufacturing environments - automotive, electronics, food processing, textiles, consumer goods. The table below shows how robotics risk varies across sub-roles, alongside a comparison with AI exposure and US wages from BLS Occupational Employment and Wage Statistics May 2024.

Sub-role scores are WorldJobsData estimates. US wages from BLS OES May 2024. Robotics risk reflects current and near-term commercial deployment in each manufacturing sector, not theoretical capability.

Why assembly is the canonical robotics use case

Assembly line robotics is not an emerging technology. Industrial arms from ABB, FANUC and KUKA have performed automotive body welding and painting at scale since the 1980s. What is new in 2026 is the extension of this capability into general assembly environments - not just purpose-built production lines, but the kind of flexible factory floor where human workers currently handle variable tasks.

The International Federation of Robotics (IFR) World Robotics 2024 report records 4.28 million industrial robots now in operation worldwide, with the global stock growing approximately 10% per year. Robot density (robots per 10,000 workers) is highest in South Korea (1,012), Singapore (730), Germany (415), Japan (419) and China (470) per IFR 2024. This density tells you where automation pressure is already highest - and where the next million factory workers will be working alongside robots rather than doing the tasks robots now handle.

The 8.5/10 robotics score for ISCO 82 reflects the maturity of this trend, not its novelty. Automotive body assembly already scores 9.5/10 - meaning the technology exists and is commercially deployed at scale. Electronics assembly at 9.0/10 is similar: Foxconn's facilities in Zhengzhou and Shenzhen, which assemble iPhones and other consumer electronics for Apple, have been deploying robotic assembly arms since 2016. The primary constraint on further automation in electronics manufacturing has been the precision required for fine-pitch components - a constraint that is narrowing with advances in robot vision systems and force feedback.

Humanoid robots: the 2025-2026 inflection

The more significant 2025-2026 development is not incremental improvements to fixed-arm robotics, but the entry of humanoid robots into factory trial deployments. Humanoid robots matter for factory workers in a way that fixed industrial arms do not, because they can operate in environments designed for human workers - using the same tools, the same conveyors, the same workstations - without requiring the physical redesign of the production environment that fixed-arm robotics demands.

Figure AI's Figure 02 robot entered a trial deployment at BMW's Spartanburg, South Carolina plant in 2025, performing body shop assembly tasks. Boston Dynamics' Atlas humanoid robot has been deployed in Hyundai manufacturing facilities. Tesla's Optimus robot is in small-scale trial production at Tesla's Gigafactories. None of these programs is yet at commercial scale, but the direction is clear: the next phase of factory automation does not require purpose-built production lines. It requires the same human worker, replaced by a robot of similar physical dimensions that can use the same infrastructure.

The 8.5/10 robotics score for ISCO 82 prices this in. The score reflects both what is already deployed (automotive, electronics) and what is entering trial deployment now. The timeline to material displacement varies by country and sub-sector: for high-robot-density countries like South Korea and Germany, it is already happening. For the majority of factory workers - in China (estimated 150M+ manufacturing workers), India, Vietnam, Bangladesh and Indonesia - the displacement timeline is longer, but the direction is the same.

Where AI does matter: quality control and scheduling

The 2.5/10 AI score for ISCO 82 reflects the minimal role of large language model AI in assembly tasks. But AI is not entirely absent from manufacturing - it appears at adjacent points in the production chain. Computer vision AI is used for quality control inspection, identifying defects in components and finished goods. AI scheduling systems optimise production sequences and materials flow. These AI applications affect factory supervisors and quality inspection workers (ISCO 31) more than line assemblers.

For assembly workers themselves, the primary impact of AI is that it enables the robotics that displaces them - smarter vision systems make robotic arms more capable in variable environments. AI is the enabler of the robotics threat, not a separate threat in its own right.

Frequently asked questions

Related analysis