The claim
The standard argument for a technology facility in an underserved area is about access: people are offline, therefore give them devices and connectivity. That argument was correct for a long time. It is now substantially out of date, and continuing to make it produces the wrong building.
The gap that remains is not connectivity. It is the class of work that cannot be done on the device people already have.
What the national numbers actually say
Sri Lanka’s 2024 computer literacy statistics record national computer literacy at 19.1%, with the Eastern Province at 10.7% and the Western Province at 31.4%[58]. That is a real and large regional gap, and it is the first half of the argument.
The second half is more interesting. In the estate sector, 92.6% of internet access is by smartphone, against 3.6% by desktop or laptop[58]. People are not offline. They are online almost entirely through a device with one small screen, no keyboard worth the name, no ports, and no ability to run professional software.
Small-cell estimates in this bulletin come with the statistics department’s own coefficient-of-variation caution. The direction of the finding is not in doubt; the exact decimal place should not be leaned on.
The phone ceiling
Here is the work a phone cannot realistically sustain. Some of it has a mobile app; none of it is practicable on a phone for hours at a time, at the quality paid work requires, or with a machine on the other end of the cable. That list is the actual specification of the building:
- Computer-aided design — modelling a part to a tolerance, which is the first step of making one.
- Running a laser cutter, a CNC router or a 3D printer.
- Wiring, probing and debugging a circuit.
- Multi-track audio recording and mixing.
- Video editing at any length that requires a timeline.
- Full-stack software development, with a terminal, a database and a browser open at once.
Every one of those is a route to paid work, and none of them is realistically open to somebody whose only computer is a phone, however good the signal is. And none of them can be solved by giving each person a laptop, because most of the list also requires a machine that costs more than a laptop and is used a few hours a week.
Shared physical infrastructure is not a compromise arrived at because individual provision was unaffordable. For this class of work it is the correct answer, and it would still be the correct answer in a wealthy country.
But it is not the machines either
Having argued for equipment, the immediate correction: a room full of equipment is not a makerspace. The defining feature is the blend of disciplines and the maker mindset, not the machines[2]. A modest rural unit that gets the culture right is a real makerspace; a well-equipped room where nobody makes anything is a storage problem with a budget line.
This is also the practical finding from research on makerspaces in resource-constrained settings, where maintainable and often low-technology tooling consistently outperforms over-specified equipment that nobody can repair[28, 26]. The cheapest technique in our curriculum is cardboard prototyping. It requires no capital, it teaches iteration faster than any machine, and it survives every budget cut — which is exactly why it is still being taught years after it stopped appearing in the written curriculum.
What follows
Three design consequences, all of which show up elsewhere on this site:
- Specify for maintenance, not for impressiveness. The Starter Unit is deliberately modest and deliberately repairable.
- Budget for people, not just for kit. The facilitator is the expensive part and the part that determines whether anything happens. Equipment without facilitation produces exactly nothing.
- Protect open access time. The value of shared infrastructure is realised when people can come and use it outside a class — which is why unstructured open making is timetabled rather than left over.