Built in a factory, finished on site - where the programme moves indoors.

MMC pays on repetition: student blocks, hotels, build-to-rent and schools where the same room or panel is made hundreds of times, and it struggles on one-off designs that change after production starts - the UK factory failures of 2022–23 proved that expensively. In the UAE the picture is different: precast concrete has been the default supply chain for villa walls and hollowcore floors for decades, so "off-site" there is business as usual rather than an experiment, and full volumetric is finding its place in hospitality and accommodation where fitting out indoors beats working on a plot at 48 °C. These eleven guides cover the whole factory end of construction - what MMC is, the systems, the DfMA discipline, the logistics, and where hybrids make sense.

Off-site construction is a transfer of work, not a deletion of it. The hours still happen - the steel is still welded, the plasterboard still fixed, the bathroom still plumbed - but they happen indoors, on a production line, with jigs, repetition and inspection at a bench instead of up a scaffold in the rain. What the site loses in weather risk and working at height, the factory gains in fixed cost: a volumetric plant or a panel line needs a pipeline of orders the way a quarry needs geology, and that commercial fact, not the technology, is the sector's real constraint. The UK factory failures of 2022 and 2023 were order-book failures, not engineering failures.

The design consequence is the discipline the sector lives or dies by: freeze early. A module's dimensions, its service routes, its door swings and its lifting points are locked when production starts, and a client change after that is not a variation but a re-manufacture. DfMA - design for manufacture and assembly - is the working method: the building designed around the factory's capabilities, tolerances analysed as a chain from casting bed to crane hook to bolt hole, and the interfaces to site-built elements designed as products, not discovered as gaps.

The site, freed from wet trades, becomes a logistics operation: just-in-time deliveries because modules do not stack in a laydown area, craneage sized for the heaviest module not the average, weather decks closed in days so follow-on trades start weeks earlier. The programme payoff is real - 20 to 50% shorter on suitable repeat products - but it is earned in the eighteen months of design and procurement before the first module ships, not in the three weeks of crane work everyone photographs.

The visitor's test for a healthy MMC project is the laydown area: there should barely be one. Modules and panels arriving just in time, lifted straight from trailer to position, with the factory's delivery schedule pinned next to the lift plan - that is the method working. A field full of wrapped modules waiting for the site to catch up is a traditional build with extra steps and a storage problem, and it happens when the factory's rhythm and the site's rhythm were never reconciled. Synchronise the two at procurement, with the delivery slot treated as firmly as the pour date, and the logistics machine runs itself.

Volumetric modules are whole rooms - or whole apartments stitched from two or three - finished in the factory to paint, sanitaryware and sometimes furniture, then stacked on site like cargo. They suit repetition with high fit-out value: student rooms, hotel keys, build-to-rent apartments, prison cells, hospital bedrooms. Their enemies are transport dimensions, double structure at every module junction, and designs that change; their friends are programme certainty and a defect rate an order below site fit-out.

Panelised systems - open or closed wall, floor and roof panels - keep the factory for what it does well, the plane surfaces, and leave the site the joints. Closed panels arrive insulated, boarded and serviced; open panels are frames the site insulates and closes. Precast concrete sits beside both: walls, stairs and hollowcore floors from the casting yard, the default structural supply chain in the Gulf for decades and a growing share of UK frames. And pods - bathrooms and kitchens as finished units - drop into otherwise conventional builds, buying the highest-complexity rooms off the critical path.

Choosing between them is a logistics and repetition calculation, not an ideology. Modules pay when the room repeats and the site is constrained; panels pay when the plan varies but the section repeats; precast pays where concrete is the structural answer anyway; pods pay nearly everywhere. Hybrid is the honest norm: a modular bedroom stack on a cast in situ core over a site-built podium is most of the successful projects you will visit.

Tolerance stack-up is the signature failure. Factory tolerance is millimetres; site tolerance is tens of millimetres; where they meet - the module landing on the slab, the panel closing on the frame - someone must have designed the adjustment in. The projects that go wrong have modules that will not sit, panels that will not close, and a grinding argument about whose tolerance was wrong. The projects that go right ran the tolerance chain at design stage, surveyed the as-built structure before shipping, and put setting-out adjustability at every interface.

Water is the second. Modules and panels travel, lift and flex; seals at module junctions, pod connections and panel joints are stressed at installation and for years after. Leaks at volumetric junctions are awkward to find and worse to fix, because the junction is buried in finished construction twice over. The defence is a tested joint design, installation by the manufacturer's own crews or trained ones, and water testing of the envelope as an assembly, not a hope.

The third is commercial: the broken pipeline. A factory with a gap in orders sheds its skilled line workers, and quality does not survive the rehiring. Clients learned to underwrite continuity - framework commitments, forward orders, sometimes taking equity in the supply chain. For the project team the practical lesson is supplier due diligence with teeth: the factory's order book, its second-source strategy for critical components, and a contract that survives the manufacturer's bad year.

In the UK, off-site is a policy ambition still finding its commercial feet: government and Build to Rent capital pushing volumetric and panelised housing, a factory sector consolidating after public failures, and building safety regulation adding a documentation load that factory traceability actually suits - a module with a birth certificate fits the golden thread better than a wall nobody photographed. The planning and warranty systems are learning to treat a factory-made house as an insurable, mortgageable object, and that institutional acceptance is now the pace-setter, not the technology.

In the UAE, precast is not a method, it is the industry: villa walls, hollowcore floors and stair flights have come off casting yards for decades, so the panelised end of MMC is business as usual with mature supply chains and pricing. Full volumetric is the novel end, finding its place where fitting out indoors beats working a plot at 48 °C - hospitality keys, labour accommodation, remote-site buildings - and where a module can be fitted out in an air-conditioned factory and shipped complete.

Climate shapes the factory case differently in each market. UK factories sell weather-avoidance and winter programme protection; Gulf factories sell heat-avoidance and quality that survives a hostile site environment. Both sell the same underlying goods - repetition, inspection at a bench, a shorter site - and both punish the same sins: unfrozen design, thin pipelines and interfaces nobody engineered. The factory moves the risk; it does not remove it, and the teams that prosper are the ones that can say precisely where it went.

Where the sector goes next is discernible from both markets: fewer heroic all-volumetric bets, more quiet hybrid competence - pods in every hotel, panels on every school, precast frames as a standard tender option, and factories plugged into frameworks that guarantee the pipeline. The technology was never the hard part; the business model was, and the industry is learning to buy continuity as deliberately as it buys components. The engineer's role in that future is unchanged: design what the factory can repeat, freeze it, respect the tolerances, and let the production line do what production lines do best.