Swedish robotics startup 120ft Factory AB has raised €576,000 from 22 angel investors to develop microfactories operated by humanoid robots. Each site is designed to fit inside three 40-foot shipping containers.
The more useful question for founders is not whether a robot can assemble a lamp. It is whether production can be purchased as a service while brands retain their existing component suppliers.
That model moves attention away from the machine and towards simulation, orchestration, inspection and the data needed to repeat a product job.
120ft’s model shows that flexible production depends on the workflow around the robot, not the robot alone.
- Paying for finished products changes the manufacturing decision from equipment ownership to production service design.
- Simulation and quality rules matter because physical automation still depends on complete product instructions.
- Founders should evaluate reusable production data before treating humanoid robots as the primary investment decision.
The factory becomes a service decision
120ft is pursuing a commercial structure in which brands do not buy its robots. They retain their existing component suppliers, send parts to a 120ft site and pay for completed assembly, inspection and packing. The company is initially targeting design-led consumer brands, including small appliance and consumer electronics makers.
This matters because the production problem described by 120ft begins before assembly. Smaller brands may need to place orders months before they know what customers will buy, while components can travel from several countries to China for assembly before finished goods return to Europe. That sequence ties cash to a forecast made before demand is clear.
What caught my attention was the separation between the component supply chain and final production. A brand can keep its existing suppliers while testing a different location and service arrangement for assembly. The trade-off is that the brand must still make its product design, parts list and quality requirements usable by the production system.
Commercial unit. Compare the price of finished output with the cost of equipment ownership, tooling, inspection and packing. The relevant decision is not simply whether a robot costs less than a worker or a production line; the material does not support that comparison.
Product fit. The model is more clearly aligned with products whose final assembly can be described and functionally tested. The lamp demonstration included assembling a cylinder lock, threading a bulb fitting and running a functional test.
Supplier continuity. Keeping existing component suppliers may reduce the scope of a first deployment, but it does not remove the need to provide complete parts and product information to the site.
The bottleneck moves into instructions
120ft’s software takes product designs, parts lists and quality requirements and converts them into digital instructions. Each job runs through a simulation before physical production, with missing information and bottlenecks intended to surface at that stage.
Once a line is running, orchestration software coordinates robots, tooling and quality checks. It also determines what happens when a task fails. That sequence is the operational core: a successful demonstration shows one path, while a production service must also define inspection, exceptions and the next approved action.
120ft says teaching one end-to-end assembly skill took seven weeks in 2025 and now takes seven hours. Its target for 2027 is seven minutes. Those figures describe the company’s reported progress on one category of assembly skill, not a general timetable for every product.
Design input. Treat the product design, parts list and quality requirements as connected production inputs. If one changes, the approved instructions and checks may need review before the next run.
Failure path. Specify what should happen after a failed task or inspection. Human review remains necessary when the defect, product or quality requirement falls outside the approved process.
Data reuse. 120ft says instructions and production data are retained so tasks can be reused and improved. The practical question for an operator is who approves those changes before they affect later products.
A robot can perform a motion, but production requires a sequence of decisions around that motion.
Microfactories make capacity portable
The physical design is deliberately compact: three 40-foot shipping containers form the basis of a site, and the combined length gives 120ft its name. Robots are shared among several customers rather than dedicated to one brand.
That arrangement changes the boundary of the manufacturing decision. A founder is no longer assessing only a machine; the assessment includes the site, tooling, software, inspection process and the way several customer jobs are handled. The source material supports the shared-customer model, but it does not establish the economics of utilisation or promise a particular production cost.
The historical pattern is familiar from cloud computing. Computing capacity became something companies could access as a service rather than purchase as dedicated hardware, but the service still depended on defined workloads, technical controls and operating decisions. 120ft applies a comparable distinction to final assembly: the brand pays for an output while the provider manages the production setup around it.
Who may benefit. Design-led brands making small appliances or consumer electronics may find the model relevant because 120ft is targeting those categories first. Their suitability still depends on whether components, assembly steps and functional checks can be specified clearly.
What remains difficult. Local assembly does not remove the need for components to reach the site. It also does not turn every product into a repeatable robot task; tooling, quality requirements and failure handling remain part of the work.
What emerges. Production data becomes more useful when instructions and results can be retained for later jobs. That creates a decision about approval and change control, not just a decision about robotics.
What is “pre seed funding”? Read the build sequence
If you are searching for “pre seed funding”, you are likely looking for how early capital supports a technically difficult product. 120ft’s €576,000 angel round from 22 investors is intended for commercial production and further development of its software and factory infrastructure. The relevant signal is the allocation: the company is building the surrounding system as well as demonstrating robot actions.
For an operator, the practical lesson is to locate the hand-off where a product design becomes a finished unit. That is where instructions, parts, tooling, inspection and packing meet, and where an automation project may create new review work if those inputs are incomplete.
Map the production record. Identify which design files, parts lists, quality requirements and test results are retained after each run. A later job can only reuse approved information that has actually been captured.
Compare service with ownership. Evaluate paying for finished products against owning equipment, while including the production steps described by the model: assembly, inspection, packing, tooling and failed-task handling.
Start with a testable product. A product with a clear assembly sequence and functional test offers a more grounded starting point than a product whose quality standard remains informal. The lamp demonstration provides an example of that type of workflow.
Keep review explicit. Approve changes to instructions, tooling and quality checks before they reach production. Robot coordination can handle defined paths, but unusual defects and changed requirements still need human judgement.
The factory is becoming a service decision only where the workflow can be specified, tested and reviewed. For more practical analysis of AI-assisted systems and automation, visit the Yanisa Execution blog and compare the production hand-offs in your own operation.
Frequently Asked Questions
How does 120ft’s microfactory model work?
Brands keep their existing component suppliers and send parts to a 120ft site. Robots support final assembly, inspection and packing, while brands pay for finished products.
How much funding has 120ft raised?
120ft has raised €576,000 from 22 angel investors. The funding is intended to support commercial production and further development of its software and factory infrastructure.
How large is a 120ft microfactory?
A 120ft site is designed to fit inside three 40-foot shipping containers. The company’s name refers to their combined length.
What products is 120ft targeting first?
120ft is initially targeting design-led consumer brands, including makers of small appliances and consumer electronics. It has demonstrated the system by assembling and functionally testing a lamp.
How quickly can 120ft teach robots an assembly skill?
120ft says one end-to-end assembly skill took seven weeks to teach in 2025 and now takes seven hours. Its target for 2027 is seven minutes, although the time may vary by product and task.