China's dark factories: so automated they don't need lights

01What a dark factory actually is

A “dark factory” is a fully automated manufacturing facility that can operate without human intervention — and therefore without lights. The term is literal: when no humans are present, there is no need for illumination, climate control for comfort, or break rooms. The factory runs 24 hours a day, 7 days a week, in the dark.

The Wall Street Journal video that anchors this article takes viewers inside Chinese manufacturing facilities where robots handle every step of production: raw material handling, machining, assembly, inspection, packaging, and even loading finished goods onto autonomous delivery vehicles. Human workers appear only for maintenance and occasional reprogramming.

The concept is not new — the term “lights-out manufacturing” has been used since the 1980s — but it has only become economically viable at scale with recent advances in industrial robotics, machine vision, and AI-driven process control. China's aggressive investment in automation infrastructure has made it the global leader in deploying these systems.

02The level of automation in Chinese manufacturing

China installs more industrial robots each year than the rest of the world combined. According to the International Federation of Robotics, China installed approximately 290,000 industrial robots in a single recent year — more than five times the number installed in Japan, the second-largest market. The WSJ video documents how this investment has transformed entire factory floors into automated production lines.

The automation extends beyond simple repetitive tasks. Modern Chinese dark factories use robotic arms equipped with machine vision for precision assembly, AI systems for quality inspection, and autonomous mobile robots for material transport between workstations. The factories are designed from the ground up for automation, with materials flowing through the production process on conveyor systems and automated guided vehicles.

The driving force behind this investment is China's shrinking workforce. The country's working-age population peaked in 2014 and has been declining since. Rising labor costs and the difficulty of recruiting workers for repetitive manufacturing jobs have pushed manufacturers to automate aggressively. Government subsidies through programs like Made in China 2025 have accelerated the trend.

03Lights-out manufacturing around the world

While China leads in scale, the concept of lights-out manufacturing is global. FANUC, the Japanese robotics giant, has operated a lights-out factory in Oshino, Japan, for over 20 years, where robots build other robots with minimal human oversight. The facility can run for up to 30 days without human intervention.

In the United States, companies like Tesla and Amazon have pushed aggressively into automated manufacturing, though with mixed results. Tesla's early attempts at full automation of Model 3 production encountered significant problems, leading Elon Musk to acknowledge that excessive automation was a mistake. The company subsequently scaled back some automated processes and reintroduced human workers in certain areas.

Germany's Industry 4.0 initiative represents a different approach: rather than full lights-out automation, it focuses on smart factories where humans and robots collaborate. Cobots (collaborative robots) work alongside humans, combining the precision and endurance of machines with the flexibility and problem-solving ability of human workers. This model may prove more resilient than pure automation for complex manufacturing.

04The impact on manufacturing jobs

The most pressing question raised by dark factories is what happens to the workers they replace. China's manufacturing sector employs over 100 million people, and the automation wave is displacing workers from assembly lines, packaging stations, and quality control roles. The WSJ video documents both the efficiency gains and the human cost.

The historical pattern of technological displacement suggests that new jobs are created even as old ones disappear, but the transition can be painful and prolonged. Automated factories require fewer workers for production but more workers for robot maintenance, programming, and system integration. The skills required are different and generally require more education and training.

In China, the government has promoted retraining programs and the shift toward higher-value manufacturing sectors. However, the sheer scale of automation deployment means that the number of workers displaced far exceeds the number of new high-skill positions created. This is a challenge facing not just China but every major manufacturing economy as automation costs continue to fall.

05Quality control by AI and machine vision

One of the critical enablers of dark factories is AI-powered quality control. Traditional quality inspection relied on human inspectors visually examining products on the line. In a dark factory, this task is performed by machine vision systems that can inspect products with greater speed, consistency, and accuracy than human workers.

Modern machine vision systems use deep learning models trained on thousands of images of acceptable and defective products. They can detect flaws that are invisible to the human eye: microscopic cracks, slight color variations, dimensional deviations measured in microns. The inspection happens at production speed, with no slowdown for quality checks.

The WSJ video shows how these systems work in practice: cameras positioned at every stage of the production line capture images of each component, and AI models make pass/fail decisions in milliseconds. Products that fail inspection are automatically diverted to a rejection bin. The data from these inspections feeds back into the production system, allowing the AI to identify patterns of defects and adjust process parameters to prevent future failures.

06The supply chain implications

Dark factories do not exist in isolation — they are part of an automated supply chain that extends from raw material suppliers to finished goods distribution. The automation of manufacturing drives automation of logistics, creating a fully integrated system with minimal human touchpoints.

Autonomous mobile robots transport materials within the factory, while automated guided vehicles move finished goods to warehouses. In the warehouse, automated storage and retrieval systems manage inventory, and robotic picking systems prepare orders for shipment. The WSJ video illustrates how a product can move from raw material to shipped package without a single human hand touching it.

This integration creates new vulnerabilities. A failure at any point in the automated chain can halt the entire system. The COVID-19 pandemic exposed the fragility of highly optimized supply chains, and dark factories concentrate this risk: when something goes wrong, the lack of human flexibility makes recovery slower. Companies are learning that some redundancy and human oversight is valuable even in highly automated systems.

07The future of human labor in manufacturing

The trajectory of dark factory technology raises fundamental questions about the future of human labor in manufacturing. As automation costs decline and capabilities increase, the economic case for human workers in routine manufacturing tasks weakens. The question is not whether automation will continue but how quickly and how society will adapt.

The most likely scenario is not total automation but a shift in the nature of manufacturing work. Humans will increasingly serve as supervisors, maintainers, and problem-solvers rather than direct producers. The number of workers needed per unit of output will continue to fall, but the value of skilled workers who can design, program, and maintain automated systems will rise.

For developing countries that have relied on low-cost manufacturing as a path to economic development, the rise of dark factories poses an existential challenge. If automation makes manufacturing cheaper in developed countries with their advanced infrastructure, the traditional model of industrialization through cheap labor may no longer work. The WSJ video closes with this sobering implication: the factory of the future may not need many of the workers who built the factories of the past.