Retrofit Factories to Fit Robots, or Build Robots to Fit Existing Factories?
Should Small-Scale Factories Adopt Robots?
Zoryor’s Industry Thoughts— Episode 1
Ⅰ.The Underlying Logic for Evaluating Robot Adoption in the Industry
The conventional industry logic goes like this: establish standardized operating procedures, complete digital systems and well-organized production line layouts first, then introduce robotic automation, and ultimately pursue benchmark lights-out, unmanned factories.
Yet this model only suits one type of facility: medium-to-large factories with mature management, highly standardized workflows, budgets for digital transformation, stable order volumes, and dedicated technical operation-and-maintenance teams. For these factories, four core criteria help decide whether to deploy robots:
1.High repetition, heavy labor and harsh working conditions: positions such as material handling, welding, grinding and palletizing, where human workers are prone to fatigue, errors and work-related injuries.
2.Stable output with infrequent process changes: high volume, low-variety production with long-term unchanged process SOPs.
3.Supporting foundational capabilities: capacity for production-line retrofits, digital systems such as PLC / MES, and on-site staff for commissioning and maintenance.
4.Calculable payback period: compare total costs of equipment procurement, retrofitting and operation-maintenance against labor costs; a payback cycle of 2-3 years is generally expected.
This evaluation framework, however, is completely unsuitable for the vast majority of small- and medium-sized factories.
Small- and medium-sized plants face real-world pain points: high-variety low-batch orders, frequent process revisions, no MES platforms, no full-time automation engineers, limited budgets, and survival-mode operations. They cannot afford to finish full scale digital construction before bringing in robots.
The industry used to build benchmark lights-out factories: hundreds of robots and end-to-end unmanned operations. A lights-out factory, nonetheless, is an outcome, not a starting point. Robots replace human labor as the final step only after factory management, digitalization and processes are fully mature. Robots cannot rescue disorganized production lines. Many small- and medium-sized enterprises blindly copy benchmark projects only to end up with idle robots, retrofitting expenses exceeding equipment costs, and total project failure.
II. A New Contradiction: Strong Demand among Small-Scale Factories without Pre-Existing Digital Infrastructure
A useful analogy: deploying robots in factories is similar to implementing AI within enterprises. Traditional AI requires high-quality data and standardized workflows to function. Conventional industrial robots work the same way: they rely on SOPs, MES and well-fixtured tooling, and fail once site conditions shift.
The market now shows a massive unmet need: countless small- and medium-sized factories lack digital infrastructure and standardized workflows, work in cluttered shop-floors and face frequent process adjustments. Still, they suffer labor shortages, difficult recruitment and rising labor costs, and they too need automation tools. This has given rise to a new category of robots: solutions that do not depend on upper-level digital systems or extreme process standardization.
Think of household robot vacuum cleaners: you do not write household-chore SOPs or renovate your entire home infrastructure.
You unbox the device, complete simple setup, and it works out-of-the-box, adapting to messy home environments.
Future industrial robots built for small and medium-sized factories should follow this philosophy:
- No large scale pre-retrofit of factory digital systems;
- No requirement for perfectly organized shop-floor conditions;
- Deployable by ordinary workers, no senior automation engineers required;
- Rapid task switching to support high-variety low-batch orders;
- Mobile and on-demand, deploy wherever work is needed.
This is the biggest opportunity brought by large AI models for industrial robotics: instead of rigid pre-programmed instructions, vision and large-model capabilities interpret real-world shop-floor conditions. Robots adapt to irregular workpieces and variable working scenarios, lowering rigid requirements for on-site standardization. The industry long overlooked such pragmatic solutions, deeming them less sophisticated in favour of large benchmark projects. Real-world development has accelerated only in the past one to two years.
III. How Should Small- and Medium-Sized Factories Evaluate These New-Generation Robots?
Benchmark standards for large scale lights out factories cannot be copied. Instead, use this more practical set of evaluation dimensions:
- Physically demanding roles with growing recruitment difficulty: jobs workers are reluctant to perform, even if processes are not perfectly standardized.
- Volatile, high-variety low-batch order mix: fixed dedicated machines are impractical; robots must switch tasks quickly.
- Inability to fund large-scale digital production line retrofits: unwillingness to invest millions upfront for MES or full production-line reconstruction.
- Acceptance of human-robot collaboration rather than immediate full automation: robots handle heavy-duty work while humans perform flexible, complex decision-making. The goal is not full site lights-out operation, but to free staff from the most strenuous positions.
- Prioritize simple deployment and low operation-maintenance barriers: operable by regular factory workers with no in-house automation specialists.
- Controllable costs and predictable payback: avoid multi-million dollar projects; adopt lightweight solutions for low-risk trial implementation.
Lights-out benchmark factories belong to mature large corporations that have survived market competition and entered profit-harvesting phases. Numerous micro and small enterprises are still fighting for survival. From the perspective of technological equity, these heavily-pressured small- and medium-sized factories are precisely the ones most in need of automation tools. They do not seek flashy unmanned factories; they want robots that work, are affordable, and can operate without extensive factory reconstruction.
The old mindset: retrofit factories to fit robots.The emerging future direction: build robots to fit existing factory conditions.
Zoryor Industry Insights