Site-Built vs Manufactured

Walk onto a conventional construction site and what you are really watching is a process of continuous adaptation. The ground is never quite level, the previous trade never quite finished to the millimetre, the delivered material varies, and the weather changes the work from one morning to the next. Yet the building comes out broadly right, and the reason is the extraordinary adaptability of skilled site trades. A bricklayer adjusts a course to take up a few millimetres of error; a plasterer floats a wall flat over a frame that is not; a carpenter scribes a skirting to a floor that dips. The site absorbs imperfection. This is not sloppiness - it is the defining virtue of assembly in place: the design can stay relatively loose about exactly how parts meet, because people of judgement will make them meet on the day.

At the heart of this adaptability are the wet trades - concrete, mortar, render, screed, plaster - materials that arrive formless and are shaped to fit whatever they are poured or spread against. A wet trade is the ultimate tolerance-absorber: it takes the shape of its container, fills the gap it is given, and hardens into place. Wet construction is forgiving precisely because nothing is pre-dimensioned; the junction is *formed*, not *fitted*. Alongside the wet trades sits craft - the cumulative skill of people who can read a situation and resolve it without a drawing for every joint. A site-built building is, in this sense, co-authored on site: the drawings set the intent, and trades complete the detail in real time.

That adaptability comes at a price, and the price is the character of the site itself. Work happens outdoors, exposed to monsoon and heat, stopping when the weather is bad. It happens at height, on scaffolding, over open edges. It happens sequentially - the foundation before the frame, the frame before the walls, the walls before the services, the services before the finishes - because each trade needs the one before it complete before it can start, and they cannot all crowd the same space at once. Quality depends on the individual crew on the individual day. Material sits in the open and is cut to fit, generating offcuts and waste. None of this makes site-building wrong; for a one-off, a tight urban infill, or a bespoke house it is often exactly right. But it explains what the factory is trying to escape - and what it gives up in return.

Now picture the same building being made in a factory, and notice what changes at the root. The work is brought indoors onto a line of workstations; the material comes to the worker at bench height under good light; jigs and fixtures hold parts in exactly the right position; the same operation is repeated many times. All of this produces consistency and speed - but it buys them with a single, non-negotiable condition: the factory works to frozen information. A jig is cut for one dimension. A line is set up for one part. A cutting list is issued for a batch. The machines and the sequence assume the drawings are correct and complete, and they will reproduce whatever those drawings say, exactly, at speed, many times over - including any error.

This is the crucial asymmetry with the site: a factory cannot improvise. There is no plasterer to float over a wall that came out eight millimetres proud, no bricklayer to take up the error in the next course. If a module is made to the wrong dimension, the wrong dimension is manufactured perfectly and repeatedly until someone stops the line. The forgiveness of the wet trade is gone, replaced by the precision - and the unforgivingness - of dry, pre-dimensioned assembly. Parts must be designed to fit *before* they are made, because they cannot be fitted *after*. Tolerance is not absorbed by craft on the day; it is designed into the joint in advance, as a deliberate gap that real, slightly-imperfect parts can still come together within.

What the factory gives in return is considerable. Because conditions are controlled, work does not stop for weather and quality is easier to hold and to inspect at the bench before the part is concealed. Because operations repeat with jigs, even less-experienced labour can produce reliable, consistent work, and the crew gets faster as it learns the product. Because the module is built off-site, its manufacture can run in parallel with groundwork and foundations on site - two phases that were sequential become simultaneous. Measured cutting and a dry store tend to cut waste; controlled benches reduce work at height and in the weather, improving safety. The factory, in short, trades the site's adaptability for consistency, speed, safety and parallel working - and that trade only pays when the design is settled early enough for the frozen information to be right.

It helps to lay the two paradigms side by side across the dimensions that actually differ, because the contrast is rarely total and almost never one-sided. Environment: the site is open, variable and weather-exposed; the factory is enclosed, controlled and stable. That favours the factory for consistency, but a factory is a fixed overhead that must be kept busy, whereas a site is set up only for the job in hand. Sequence: site work is largely sequential, one trade after another; off-site allows parallel working, with modules built while the ground is prepared. Parallelism is the single biggest programme lever prefab offers - but it is realised only if the design is frozen early enough for the factory to start.

Quality and consistency: bench work, jigs and inspection tend to make factory output more consistent and easier to check before it is concealed; site quality is more variable, depending on the crew and the day, though a superb crew on a good day can match anything. Labour: the site relies on a larger number of mobile trades moving through; the factory uses a smaller, more stable workforce doing repetitive operations, often less reliant on scarce high-skill trades. In much of the West, where site labour is expensive and scarce, this is a primary driver for going off-site; in India, where construction labour is abundant and comparatively inexpensive, this particular argument is weaker, and the case for prefab rests more on scale, speed, repetition and quality.

Waste: controlled cutting, optimised nesting and a dry store tend to reduce material waste in the factory; open-site cutting and weather exposure tend to increase it. Safety: moving work off ladders and scaffolds to a bench, out of the weather and off a congested site, generally improves safety - one of the least-disputed advantages of off-site. Change and flexibility: this is where the site wins decisively. A site-built project can absorb a late change - move a wall, swap a finish, adjust to a surprise found in the ground - relatively cheaply, because so much is still being decided as the work proceeds. A manufactured project cannot: once jigs are cut and modules are in production, change is slow and expensive, and a change discovered after modules are built may mean scrapping them. Flexibility and adaptability are the price the factory pays for its consistency and speed.

The honest conclusion - the one this whole course insists on - is that neither paradigm is universally better; they reward different project shapes. Site-built construction rewards the one-off, the bespoke, the tight or awkward site, the project where the design will keep evolving or where the ground holds surprises, and the context where adaptable labour is plentiful and cheap. Its adaptability is a genuine engineering virtue: the ability to absorb error and accommodate late decisions has real value, and for a great deal of building it remains the right and sometimes the only sensible approach. To treat site construction as backward is to misunderstand what its flexibility buys.

Manufactured construction rewards the opposite profile: repetition (many similar units - hotel rooms, flats, wards, classrooms), volume (enough output to keep a factory busy and to climb the learning curve), early certainty (a design that can be frozen before manufacture starts), a programme under pressure where parallel working pays, and a context where consistent quality and safety are worth paying for. Where those conditions hold, the factory's discipline turns into real advantage. Where they do not - a unique one-off, a design still in flux, a project too small to absorb the factory overhead - forcing it off-site tends to cost more and deliver less, and the adaptability you gave up is exactly what you needed.

This is why the site-built-versus-manufactured choice is not a loyalty test or a technology fashion; it is a judgement about fit, made early, based on the shape of the specific project. And it is rarely all-or-nothing: most real buildings sit somewhere between the poles, mixing a site-built frame with panelised walls, or dropping factory-made bathroom pods into an otherwise conventional block. The next lesson opens up that middle ground as a spectrum. For now, carry the core contrast: the site is a place of adaptation, where trades and wet work absorb error and late change at the cost of weather, sequence and variability; the factory is a place of precision, where frozen information and repetition buy consistency, speed, safety and parallel working at the cost of flexibility. Knowing which virtue a project actually needs - adaptability or consistency - is the beginning of using either one well, with the binding structural, fire, acoustic and connection engineering, and whether the business case truly stands, left to the qualified specialists, the manufacturer's tested system and the governing codes.