Investigating Building Production Times

Vehicles 

So from previous, a vehicle assembly line can produce 100,000 vehicles per year, and vehicles can exit the line at the rate of 1 vehicle per minute or 4 minutes. The line operates continuously, and some lines can produce more each year. The total production time used to be indicated as between 2 and 4 weeks, though now appears to be 18 to 35 hours so less than 1 week. However, apparently can expect it to take a few months when order a vehicle, even though there are acres of land covered with unsold vehicles. So whilst the vehicles can be produced quickly, the right vehicles are not necessarily getting made in the right places.

If everything was right, then the next vehicle of the production line a minute from now would be yours. However, if want custom features, then your vehicle has to flow from the start of the production line to the end acquiring all the custom features. So if your vehicle can be put on the line immediately, then will have to wait 18 to 35 hours. But if there is a waiting list or orders to be filled, then it could takes weeks before your order reaches the production line. What really want is, to be able to go to the car sales lot, pick a car and drive away the same day. So cars being sold are not necessarily the cars which are desired, they are the cars which are available, which can otherwise fulfil the dependency on a car.

{Global market apparently around 57.5 million units each year}

Shipping Containers

The largest manufacturer of shipping containers produces 2 million containers each year in 11 factories, so averaging 181,818 containers per factory. 

{Global Market apparently 1.04 billion twenty foot equivalent units (TEU) each year}

Housing

The housing crisis is global not just a local issue. A housing shortage has existed for decades. Currently there is an estimated 1.6 billion people lacking adequate housing, and expected to grow to 3 billion by 2030. Back in the 19080's there was also an indication that 75% of global housing was inadequate, I'm not aware that this has been changed. One prediction is that maximum sustainable population is 10 billion. The world reached a population of 8 billion in 2022, so we are getting close to the limit. However, the limit is still greater than the current population, and only an estimate, so still useful for planning. Whilst some countries are experiencing population decline, the world population is still growing.  So assuming an industrial city-state of 100km diameter, and maximum population of 10 million, then 1000 such cities will be needed. Assuming population distribution matches current distribution, then each country has more cities than needed, and population can be distributed to smaller cities over larger area.

If consider most of the population is going to live in sole occupancy units, then need 10 billion units, if population to be coupled up then need 5 billion units, if families of 4 then will need 2.5 billion units. Given average household occupancy of less than 3 persons, then will need 3.3 billion units.

Given that average household occupancy is less than 3 persons and we have a housing shortage, the implication is that average household occupancy needs to be 1 or 2 persons. So assuming in the next five years need to produce housing for 3 billion, then 0.6 billion each year sole occupancy units. A sole occupancy unit is typically adequate for a couple, and may just be tolerable for a young family. In any case if aim for modular sole occupancy units they can be combined to create family units. The number of modules basically remains the same, though can possibly save a few modules on family units compared to multiple sole occupancy units, as facilities like kitchens, bathrooms and laundries can be shared. 

If only make 0.6 billion units each year, then maybe considered as not adequately addressing the current shortfall. Should want to see currently shortfall removed as quickly as possible, not a 5 year plan, or 15 year plan and certainly not a 30 year plan. Shortfalls should be eliminated in less than 1 year. So that would be 1.6 billion sole occupancy units in 1 year, or 0.4 billion family units. which would then leave 1.4 billion to be provided in remaining 4 years, so drops to 0.35 billion sole occupancy units each year.

Assuming 99% of buildings already exist, then possibly only need to cater for the shortfall. However, if 75% inadequate also need improvements. The main inadequacies are kitchens, bathrooms, and laundries, that is water supply and sanitary drainage. Here in Australia, expect around 5% of households to undertake renovation involving bathrooms, kitchens, verandahs, or outdoor decks, or possibly adding an extra bedroom.

So if have a city with zero population growth, and it has a population of 10 million, in 5 million dwellings. Then estimate a need for production of around 250 thousand units every year for replacement and maintenance. Which if a single factory can produce 100,000 units, would need at least 3 factories to each city.

Or from another perspective Australia with a population around 25 million, only needs 3 industrial city-states, and therefore a total of 9 factories. It has more cities and potential for more factories. If it has more factories then it can produce more units, and if it has more units available, then more dwellings can be  renovated or replace each year.

Modular Buildings

To be clear offsite prefabricated does not equate to modular. If have true modular then floor plans are determined from the available modules. If don't like the floor plan after a few weeks, can rearrange the modules, if want more floor area can acquire more modules, if want to increase the number of storeys can again add more modules. If at some point in the future want less space, then can remove and sell the modules which are surplus to current requirements.

If simply have offsite prefabrication, then floor plan is fixed, and future renovations are hindered. As simply offsite prefab and custom designed, new customised  transportable segments have to be designed and fabricated to fit the existing segments. It is time consuming and expensive, compared to buying off-the-shelf modules. It is possible that off-the-shelf volumetric modules can be delivered within 24 hours, and within a day the house is extended. No waiting for design, approvals or fabrication.

Production of Prefabricated Buildings

Mean while manufactures of modular buildings are only able to produce a few thousand (ie. 4000) modules per year or tiny house builders and transportable house builders are only producing around a few hundred (ie. 250) each year. Time frames for supply are also suggested at around 4 to 12 weeks, or longer at 18 months. 

Is the short fall for builders a consequence of process, a technical limitation or a consequence of low demand?

There is a global shortage of houses, and prefab buildings have to be designed as transportable segments or units. These segments/units have to be constrained to the same dimensions as vehicles, this puts them into the same size range as cars and shipping containers. The buildings can be shipped to the world. It would appear that the demand far exceeds the willingness to supply.

From previous, a box from composite panels can be produced in 10 minutes, a caravan with kitchen and washroom in 7 hours, and the shell for a larger building in one day (say 8 hours). A tiny home can be framed in cold-formed steel in 4 hours.

Assuming the following production stages:

1. Floor Assembly
2. Wall Assembly
3. Roof Assembly
4. External Roof Cladding
5. Electrical Systems (1st)
6. Plumbing Systems (1st)
7. External Wall Cladding
8. Internal Wall Linings
9. Internal Ceiling Linings
10. Electrical Systems (2nd)
11. Plumbing Systems (2nd)
12. Windows and Doors
13. Internal Decoration
14. Bathroom Fitout
15. Kitchen Fitout
16. Quality Checks

and 1 day at most for each stage, then expect a upper limit of 16 days., so at 5 work days per calendar week, then needs around 4 weeks to produce. But clearly the caravan industry can complete all these stages in one day, and a typical module is not expected to be larger than a caravan, furthermore a typical module is empty space. So just looking at the shell:

1. Floor Assembly
2. Wall Assembly
3. Roof Assembly
4. External Roof Cladding
5. External Wall Cladding
6. Windows and Doors
7. Quality Checks

This appears achievable in 1 day, leaving:

1. Electrical Systems (1st)
2. Plumbing Systems (1st)
3. Internal Wall Linings
4. Internal Ceiling Linings
5. Electrical Systems (2nd)
6. Plumbing Systems (2nd)
7. Internal Decoration
8. Bathroom Fitout
9. Kitchen Fitout
10. Quality Checks

Now if the electrical and plumbing were integral with floor, wall, and ceiling panels, along with internal linings. Then remaining stages would be reduced to:

1. Internal Decoration
2. Bathroom Fitout
3. Kitchen Fitout
4. Quality Checks

The vast majority of modules are empty space, and don't need bathrooms, kitchens or laundries, so that just leaves internal decoration and final quality checks. So if we make wet areas as separate modules, then we can speed up the provision of basic shelter. Wet area modules can be produced on a separate production line, and then they are plugged into basic space modules to create a whole dwelling. As modules for wet areas are smaller then the shell can be produced faster. The rooms possibly expect in a dwelling with recommended dimensions for resale are:

1. Kitchen (3m x 2.7m.)
2. Bathroom (1.8m x 2.4m) /WC (1.8m x 1m)
3. Laundry (2.7m x 2.1m)
4. Dining Room (3m x 3m)
5. Lounge (6m x 3.6m)/Living Room
6. Master Bedroom (3.6m x 3.6m)
7. Additional Bedrooms (3m x 2.7m)

In South Australia, minimum housing standards require a room to have a minimum area of 7.5 sq.m to be considered a bedroom (ie. 2.74m x 2.74m). Also note that these dimensions are internal usable space, however most of the dimensions are recommendations not requirements. Also broad loom carpet is 3.6m, so anything wider and the carpet has to be cut and has a seam.

So if we have 3.6m wide rooms and a 1.2m wide corridor between then get a maximum width of around 8.4m, allow 0.3m for external walls and 0.1m for internal, then add an extra  2*0.3+2*0.1=0.8m bringing over all width to 9.2m. So don't expect to span further than 9.2m: which wouldn't need to do as have modules providing walls at 3.6m centres or less.

Possible modules:

1. Narrow (2.4m high x 0.6m wide x maximum 12m long)
2. Small (2.4m high x 0.9m wide x maximum 12m long)
3. Basic (2.4m high x 1.2m wide x maximum 12m long)
4. Wide (2.4m high x 2.4m wide x maximum 12m long)

These would be internal dimensions, so assuming external walls do not need to be more than 150mm thick, then over all width of wide module would be 2.7m and not readily transportable. Though the basic module wouldn't need walls, and these can be transported separately and installed on site, in which case the thickness of the walls can increase to 300mm or so each.

So to create a lounge or bedroom would need either 3 basic modules side by side, or one wide module and a basic module, to get width of 3.6m. Modules can be made with two opposite walls missing, or open on one side only. So a 2.4m wide module never has side walls as it would make it too wide to transport, whilst the smaller modules do have one or both the side walls as they wouldn't be too wide to transport. Also the smaller modules can be defined by over all width, or internal width only.

So assuming 200mm insulated wall to one side only, then a 600mm wide module can be considered as providing an additional 400mm of internal space. So fitted to each side of a 2.4m module provides an additional 800mm of internal width, to get 3.2m of internal space.

To create the 8.4m requirement, as 2 segments 3.6m and central corridor 1.2m wide, can use 2 wide modules and 3 basic modules. One basic module, the corridor, can be provided with walls on both sides to provide internal division into rooms. The other basic modules can either be combined together, or placed on the exterior of each room. If placed to the exterior then these modules can have the external walls fitted.

Then the other approach considers that 8.4m is less than 12m, so can place 2.4m x 8.4m modules end to end to create a building 8.4m wide of any length. So when comes to basic rooms and empty space the modules are not such a problem. 

Creating rooms which have to be fitted out before transported can be a problem, as may not be possible to create a room large enough which can be transported. Creating a WC seems viable, whilst a bathroom at 1.8m x 2.4m is viable if adopt a 1.8m wide module and as long as needed. A laundry at 2.1m x 2.7m is relatively large, and probably more often about 1.8m wide. This allows 900mm width for equipment on one side and a 900mm walk through to the back door. So again a 1.8m wide module as long as needed, or alternatively adopt 2.4m wide modules. That primarily leaves  the kitchen as a problem, it being 3m x 2.7m, with neither dimension suitable for transportation. But then again tend to have cupboards going along one wall, or opposite walls creating an aisle between. So could use a 2.4m module and a 0.6m module, or two modules 1.5m. Or module could be 3m wide on open faces, and then 2.4m wide strip, which are placed end to end to create any length.

So to create longest buildings, do not span the 2.4m maximum width of segment, instead span the 12m maximum length of segment and add 2.4m segments end to end to create buildings of any length. In general wouldn't expect a need to span more than 9.2m. These segments can then have internal walls to create rooms from multiple segments.

So assuming that an empty shell can be completed every day, and that if the shell is assembled from 6 composite panels it takes a 10 minutes to assemble. Then in a 480 minute day, assuming an adequate supply of panels, can produce 48 modules each day. So with 250 productive days, 12,000 modules each year. If have 3 shifts then 36,000 each year. If operate continuously with as many shifts as needed, then can produce 365.25*24*60/10 = 52,596 units each year. 

But if split the task into stations, then assume have 5 panels to attach to a base, so have 5 stations and each takes 2 minutes. So flow along line is at rate of 1 completion every 2 minutes. So in 24*60/2=720 units each day, and for the year 720*365.25 = 262,980 units.

If a sole occupancy unit requires 2 modules, then can produce 131,490 dwellings each year, and 2 factories would be required to produce 250 thousand dwellings each year. This is based on an assumption that takes 2 minutes to install a panel: it likely takes longer than that simply to move panel into position.

If the panels are light enough then an empty box can be fabricated relatively quickly. The panels have to be fabricated away from the main assembly line, and they have to be produced relatively quickly. So assuming 12m long, 2.4m high and 2.4m wide, and basic panel is 1.2m wide. Then need to produce 2 floor panels and 2 roof panels each 12m long, so total of 48m of panel. Similarly 48m for side walls, and 4.8*2=9.6m for end walls. So total of 105.6m of panel. {Alternative 10 panels x 1.2m wide x 2.4m high gives 24m each side wall, so 48m for both side walls.} Assuming panel formed at the rate of 2m per minute, then it will take 52.8 minutes to produce, which is slower than the building is assembled. Stock piling doesn't help, as will still eventually run out off panel, and have to wait. Therefore need more than one machine producing panels in parallel. Alternatively accept a lower production rate for the boxes and a need for additional factories or building production lines.

Also note our bench mark is 10 minutes to assemble a composite panel box, and 30 minutes to assemble cold-formed steel framing. The latter than needs to be clad both externally and internally, pushing production times still higher. But should still be talking about production times in minutes, not days or weeks.

A bathroom or kitchen renovation is likely to take 1 or 2 days, a week at most once they get on site. Most of the cupboards are fabricated offsite in any case, then adjusted to fit on site if necessary. It takes more than a day on site then really need to know what they are doing. So I don't expect wet area modules to take more than 1 to 2 days to produce.  Also expect that most of the tasks involved would be less than 2 hours duration, so have potential flow at rate of 1 unit every 2 hours. So production of 12 modules each day, and 3000 modules each year, for 250 productive days.

Also note that depending on nature of building if less than 15 sqm here in SA, then may not need development approval. A factory can build laundry buildings (5.67 sqm) all day every day. Also a 6m shipping container is 14.4 sqm. Need extra bathroom (4.2sqm) space for the kids, then install a module, which can be accessed from the back laundry door. The laundry walk through is extended and the back door relocated, so don't have to walk outside to access the new bathroom. When no longer need the extra bathroom facilities then get rid of the module.

So modules are not simply for constructing new dwellings, they are useful in own right for extending existing dwellings. The modules have to be readily available, faster supply, and lower price than onsite construction. Simply shifting on site processes off site into a factory offers little benefit. There has to be concurrent design of product and process, and process has achieve consistent output so that have higher quality in a factory than on site. Simply shifting into a factory does not improve quality, nor productivity.

A single work team, with appropriately scheduled work processes, should be able to produce at least 250 modules each year: assuming no task takes longer than 1 day. If tasks can be further reduced than greater production can be achieved. 

Note that whilst work can be staged, and crews can move between construction sites, such travelling is a waste of time. By shifting into a factory, only one site crew needs to travel to the site, most workers only travel to the factory and along a production line. A lot of subcontractors or trades will charge for a days work, even though only a few hours or even a few minutes work, and this is because getting to the site wastes the day. In a factory they can more readily move from building to building, and therefore can complete more buildings in a day, than they can if working on site. Therefore the price of factory built buildings should be lower. However have extra costs of transporting large object, and possible crane fees. But have these transportation and crane costs for large fabricated components. If building can be delivered by tilt trailer then transportation costs should be low.

So need more process time information, so that can move away from estimates based on hours and days, to estimates based on minutes, and get the flow rates better. So not interested in a day to fitout a bathroom or kitchen, really interested in how that time is built up, and if its possible to get such modules flowing from a production line at so many units per minute.

---

Related Posts:

...

---

Revisions:

[04/01/2025] : Original

Comparing Production Times

Background 

As a structural design (engineering) consultant I rarely get to see fabrication and construction processes, this is because most designs are "standard designs" for manufacturers, so they could be built at any time. If I do get to see the end-product, it is typically illegal as-built-construction, which needs to be assessed to avoid demolition. So the only real access to fabrication and construction processes I have at the present are youtube videos of processes posted by manufacturers and builders around the world for promotional purposes. Such videos are not entirely suitable for work study and work measurement, but they do provide some insight.

South Australia Housing Crisis

At the beginning of the 1990's here in South Australia, average household occupancy was less than 3 persons per household and vehicle ownership at 2 vehicles per household. We had 2 vehicle assembly plants just before, giving a total capacity of 180,000 vehicles per year, this reduced to 1 assembly plant at 100,000 vehicles per year. On average had 1 person per household not yet coupled, with no car and no house. To couple up the population and reduce occupancy to 2 persons per household would require 1.5 times the housing stock, and land. The vehicle industry could provide these couples with private space in the form of a single vehicle in less than 3 years, assuming all production used for one purpose. In less than 5 years the couples would have 2 vehicles. The building industry was producing less than 10,000 dwellings each year. So they should have provided the couples with housing by the year 2015.

Here we are nearing 2025, and average household occupancy still less than 3 persons per household but not yet 2. The Adelaide metropolitan area has grown to the edge of rural towns, large areas of open space have disappeared under housing. We no longer have vehicle assembly plants, we have a housing crisis, and shortage of skilled labour in the building industry.

From my view there has always been a housing shortage and the building industry been unable to supply. The housing we have is not suitable, and is difficult to adapt. Most of the existing housing is single storey, 3 bedroom, brick veneer on timber frame, with concrete floor slab on ground. Large numbers of these houses are occupied by one to two persons, not by families.

Always Been Housing Shortage

Also when we experience bushfires, floods and tropical cyclones, it takes far too long to restore housing. Modular housing could be produced faster, and provide more adaptable buildings. However to be fully adaptable also need to be able to change the division of land: that is the division of land has to be more adaptable more rapidly. That is larger blocks with internal division into sites, sites which can vary at an instance.

Modular Production

However, there seems to be a problem with production. Both established and new players in the modular building space, do not appear able to produce more than around 4000 modules each year. Given that a building is largely empty space, and modules to be transportable have to comply with the dimensional constraints of vehicles, it should be possible to produce such modules as fast as or faster than vehicles. Also the largest shipping container manufacturer in the world produces 2 million containers in 11 factories, so an average of 181,818 containers per factory.

It has been suggested that the market isn't large enough, they need more volume on the demand side. That may be true, but again an average of 2 vehicles per household, so if building construction at rate of 10,000 then vehicle assembly plant only needs to be producing 20,000? The vehicle assembly plant is 5 times larger than needs to be, no wonder it closed down. Except the vehicle assembly plant was exporting, and also producing replacement vehicles. But cannot export buildings, and buildings last 50 years or more. But can export building materials, and vehicle sized building modules, also vast majority of building activity is renovating existing construction and renovations can be done using modules.

It should be possible to produce modules faster, and modules can be produced as flat pack kits. Can transport a lot more flat pack kits over seas than empty boxes. But then the box doesn't have to be empty, it can be filled with furniture and appliances, and so used as a storage container. Want to go over seas, then don't need to pack all possessions into a container ready for shipment, just pack the whole house up and ship the whole thing.

Global Housing Shortage

So there is a global housing shortage, and skilled workers don't necessarily have any desire to go live in a tent whilst building houses in a developing country. So moving skilled workers to where needed a potential problem. But if can transport the buildings then the workers and factory can be located any where. Cars are expensive, and houses more expensive. But why is the empty box more expensive than the complex machine? Excluding the cost of land: as who wants to own land which is merely space: it doesn't have natural water supply and doesn't produce food, and if government or community wants to put highway through the land or submerge below water then they will.

When I was in school, there were people who had a mortgage on a house in Adelaide, but were working and renting in Sydney or Melbourne. They couldn't sell their houses because there was a lack of employment in Adelaide. If just rented land, and houses were transportable, then could just move house to new location. When move to new locality typically stuck with the houses which exist, and these are not always suitable, and not always viable to make suitable. Land is wasted because unsuitable houses get built. If houses can be relocated, then houses which better suit can be moved in and out of neighbourhoods. Whilst modular houses can expand and contract as needed.

But first need to see if can produce faster. If take the 4000 units a year as a bench mark, the question is does this require 25 production lines to produce 100,000 units per year or is the process capable with out modification and the suppliers merely have low demand?

Also important to understand that whilst a vehicle assembly plant suggests that produces 1 vehicle every 4 minutes, 2 minutes or 16 seconds, this does not mean it takes such short amount of time to produce a vehicle. It likely takes around 2 to 4 weeks to fully assemble a vehicle, however if they are to flow from the line at the rate of 1 every 2 minutes, then no operation can take more then 2 minutes. So if have a 4 minute task on the line, the flow would be one every 4 minutes, it doesn't matter if other tasks take a few seconds, and it doesn't matter if the last task only takes a second, product cannot exit the line faster than 1 every 4 minutes. There will be waiting along the line.

The flow is similar to water in a pipe. It may take 2 weeks for water to flow along a pipe from the reservoir to your house, but once there it flows from the pipe at a rate of around 10 litres/minute. But it cannot do that until the pipe is full. The same is for the assembly line, once the line is full of partially assembled vehicles then completed vehicles can start to flow from the end of the line.

Need for Distributors

Another point of note is that the only reason vehicle manufacturers can sell near to their production quantity is because they have distributors around the country and around the world. Currently they do have a problem and there are large areas of land covered with vehicles which cannot sell or at least have not sold. The change over to electric vehicles is a partial problem, and existing product not being suitable for the remaining demand. With an average of 2 vehicles per household and 3 persons per household, it is unlikely that all 3  possible vehicles need to be oversized, fuel guzzling, 4 wheel drive, all terrain vehicles to travel 1km to 5km to local stores. I will hazard a guess that as electric becomes more viable, that there will be additional vehicles per household, and that they most likely will be electric bikes or trikes. Further more that smaller vehicles will be capable of being coupled to form larger vehicles. The vehicles will also be more adaptable, so that enclosed in winter and open in the summer. Electric vehicles have less constraint on shape and form compared to vehicles with internal combustion engines (ICE). Whilst battery packs and electric motors impose some constraints on form it is not as restrictive as the ICE and its supporting machinery. There is also the issue of autonomous robotic drones which can travel to the stores and retrieve goods on their own, these will occupy even less space. So private mechanised transport is going to change, and so existing manufacturers will have some transition problems as they transform their factories or choose not to change, and continue to supply for the reduced demand.

So there are problems with factories and substitute products. However, the problem with buildings is onsite versus offsite fabrication, and whether a box is empty or fully fitted out. The largest structures tend to be assembled from components made in a factory: large structures do not merely consist of buildings and bridges, there are also ships and aircraft and various other utility structures.

So need to get the small to medium sized buildings off-site. Identify the issues keeping the fabrication and construction on-site, and keeping buildings being anchored to the ground by methods which make it difficult to relocate the building.

So what timing can we expect for production of a building module?

Production Examples

Cold-formed Steel Framing

So here is a video which indicates that takes 30 minutes to assemble the panels of a tiny home onto a trailer, the framing produced by Dynamic Steel Frame.

Here's another showing more activity:

Whilst Dynamic Steel Frame are mostly cold-formed steel (CFS) building fabricators, they do have a few videos, of tiny homes, modules and pods. On a LinkedIn post they indicated a time of around 4 hours to roll form the steel framing and assemble the framing. So subtracting the 30 minute assembly time, the roll-forming of the sections and assembly of the panels takes around 3.5 hours. Check out their other videos and buildings here.

So CFS provides a relatively quick means of providing framing for a small module.

Timber Framing

Here are some examples of more complete construction of timber framed building, constructed on site.

The building is 14 ft x 28 ft (4.2m x 8.5m), so larger than a tiny home on wheels (THoW) which typically restricted to 2.5m wide. So this shed building is assembled and clad in one day.

And another example by the same crew Atlas Backyard Sheds, this time 16ft x 40ft (4.8m x 12m), again larger than a THoW, and I believe completed in less than 1 day.

There is additional work required internally, but we have an indication that the shell for a building can be constructed in less than 1 day.

Container

Now here is an example of the assembly of a simple 16ft (4.8m) box from composite panels. The panels are the full size of the box, and now space is enclosed in less than 10 minutes. {Assuming video is real time.}  The associated company appears to be a producer of the composite panels: HolyCore.

So if use full panels for each side then can significantly reduce time to assembly a box.

Caravans

Here is a caravaners tour of the Bailey caravan factory.  In part 1, can see the composite panels which form the shell being shaped.

and in part 2, see the components assembled into the caravan. Also note it seems a relatively common practice to keep the walls and roof off, whilst fitting out the interior. This means do not have to work in a cramped or dark space.

Here's a video from Bailey themselves, from which we get it takes an average of 6 hours and 48 minutes to make a caravan and over 12,000 components for each caravan. The caravan being fabricated is indicated as the Bailey Phoenix+ 640, which has berths to sleep 4.

Here's a tour of Adria Mobile factory providing another view of the equipment available to improve production, this time motorhomes not just caravans.

Truck Bodies

Still another comparison is that of truck bodies, these are similar to light weight shipping containers, with all sides of the box being full size: not an assembly of smaller panels.

Manufactured Homes and/or Static Caravans

In the UK they have static caravans, whilst in the USA they have HUD specification manufactured homes. Here in Australia we just have transportable houses designed and built to the same code as housing. Though the tiny house movement has arrived to confuse issues, such buildings are considered caravans if on wheels, otherwise they are considered to be houses and have to comply with building code.

Here is an example of static caravan production by Swift:

And here is a tour of a manufactured home factory by Skyline Homes presented by Factory Home Expo Centers

Not much different than a site built home of light weight construction, here in Australia, except here we don't need to provide a permanent wheel chassis. In Australia we normally transport , relocatable or transportable houses on the back of a truck, it doesn't need its own wheel set. When it needs to be moved the building gets put on the back of another truck.

Here is the start of a series of videos on manufactured homes by Oakwood Homes Tulsa.

Then there is Boxabl who have built a massive factory, but still only down around 4000 units each year.

Prefabricated or Segmented Transportable

At the moment few manufacturers claiming to be modular are truly modular. The majority simply take a floor plan and divide into transportable segments and then set about designing and building each segment. For true modular, the floor plan is determined from the available modules. If the floor plan determines the modules then have segmental transportable not modular.  As long as Boxable does not customise its floor plans and buildings, then it is modular.

An example of prefabricated, with walls having appearance of brick, by TopHat Homes. It isn't necessary to loose the appearance of brick, nor is it necessary to have the appearance with all the disadvantages of brittle bricks. Important here in South Australia, where mostly have brick veneer and large footings to minimise cracking due to soil heave. With masonry panels can have the decoration without the expense.

An example of robotic production for walls, by Autovol, the question is why don't they have a more continuous production of the walls, compared to the wall segments?

Here is something which appears to be modular, as long as the modules do not change from project to project, then can consider them to be true modular, otherwise they are just prefabricating transportable segments.

Some History

A Nissan Hut isn't entirely prefabricated but it is a simple kit to assemble.

Early prefab steel house:

and mail order kit homes:

and post war reconstruction in the UK:

And some historical background to prefabricated:

Prefabricated the Way Forward

I believe that prefabricated will be the way forward. There has always been a housing shortage, the current crisis is not so much locally generated, but generated by migration. People have legs, they are not plants, they are meant to be mobile. The problem is when they settle, put down roots and stay.

There was an idea to allow migration to create pressure for housing to keep the building industry going. The consequence, average household occupancy did not reduce to 2 persons per household it stayed close to 3 persons. So we covered roughly 1.5 the area of land to get additional houses and we need to do so again, and talking about immigration to provide the labour force. But same problem exists across the industrialised world. So where would the skilled labour come from, and when they arrive where will they live? Competition in the labour market, and attracting skills away from one place to another, doesn't solve the problem.

Buildings can be produced in a factory, they can be made as flat packs or used as boxes filled with packaged household contents. So ship can transport more houses than cars.

So buildings can be produced in a factory, the buildings can then be transported over seas, to the locations the migrants are coming from. The migrants cannot be coming from the industrialised world, as they all have the same problem. So if the trades migrate, they increase  the shortage at their point of origin, and reduce it at their destination: but over all the shortage remains. If migrants come from the developing countries, then those countries loose the resources they need to develop.

Houses are no longer made from local materials, all materials used are transported relatively large distances using mechanised transport. So unless making adobe bricks on site, and have a nearby woods the materials are going to be transported to site.

In the industrialised nations fertility rate is less than the 2.1 required for replacement. This is potentially good, as should not want further growth of our cities, drawing resources from ever more distant locations to feed an ever increasing population. So migration could be constrained to balancing a need for replacement. However in some situations, shrinking the cities back to previous size may be preferable, in which case halt migration altogether, allow the city to shrink, demolish the buildings and reclaim the open space.

Other issue is why do people migrate. Typically its some improvement in lifestyle and economic gain, in which case it suggests a need to modify something at the point of origin to encourage people to stay. For the developing and undeveloped countries the shortfall is the supply of food, clothing and shelter, and provision of education and health care. Furthermore migration is occurring in these countries from rural areas into cities. So whilst population maybe created in the rural areas it is flowing into the cities.

As people who remain in the rural areas age, there comes a time when there is a problem getting people to live and work in the rural areas. People are needed in the rural areas to to produce food, fibre and fuel. Living conditions in the rural towns needs to be improved and that largely requires improvements in the logistics of supplying the rural regions.

Rural towns don't necessarily need permanent facilities. Mining towns need to be mobile, even if there for a hundred years. Towns thus can move around in remote regions, and so can various service centres. If a town and its services are mobile, then it can adjust its location relative to other more residential towns. A service town may start out closer to one rural community than others, then it adjusts its location so that it is equal distances from many rural communities. Also people don't have to travel to service centres, the service providers can travel to the customers. So the services themselves are mobile.

It is not necessary for all buildings to be permanently anchored to the ground, and certainly not our residences if we need to relocate for continued employment. Established buildings are a problem as they seldom meet current needs. Consider from an alternative perspective your need is for a mini moke and the only vehicle available is a Mack truck. Similarly if need a small apartment, and only house available is a large multi-bedroom house with large gardens. Not only is the available house over sized there is a lot of work keeping the house and gardens maintained. The available product is unsuitable. 

Another point is that caravan sites provide utility poles or boxes. There is no reason why residential sites do not provide like wise for housing. All services come into the site and need meter boxes, these can all be collected in the one place and sometimes are. A utility service box can be provided at the front of the site, and underground service access channel provided to reach the building. The service channel remains accessible. It is not an earth covered trench with a drive way over it. The whole system is then designed so that the building can be readily connected and disconnected from the services.

Also note that caravans can be fully self-contained, and do not require a permanent connection to services. They do need to dispose of waste and replenish resources. This is important as cities are stretching beyond their available infrastructure, so caravan technology is a viable solution in the short term. Also there is no point building additional infrastructure if population spike is temporary.

So majority of houses are 3 bedroom, and average household occupancy is less than 3 persons. Therefore each existing site is potentially designed for 4 persons, comprising of 2 adults and 2 children. There is no apparent need to release more land as residential, rather a need to modify a large number of existing dwellings to better suit the needs of sole occupancy or two person occupancy. To keep people in their neighbourhoods and using the existing infrastructure.

The main functional rooms in a house are kitchen, bathroom and laundry and these can be shared resources. All other rooms are largely empty space. Also minimum site area in a caravan park is 100 sq.m, or 10m x 10m, it used to be 81 sq.m or 9m x 9m. On such a block it is possible to place a 6m x 6m sole occupancy unit, which may be equally suitable for a couple.

So houses on large residential allotments with 100 sq.m to spare have the potential to provide ancillary accommodation. This has the potential to double the number of dwellings, reducing household occupancy to approximately 1.5 persons per dwelling. Also in first instance a module 2.4m x 6m or 2.4m x 4.8m is likely to be adequate, thus a smaller site would be suitable.

Also modules can be made to match existing, by demolishing part of existing and salvaging the bricks, then cutting those bricks into tiles, and make wall panels using the tiles. For example if only cut the bricks in half then can cover twice the surface area. The bricks are 110mm thick if cut into 10mm thick tiles, then get 11 tiles from each brick, so even more surface area covered. Other materials are required for the wall assembly, but over all for a given thickness of wall, the wall has better performance.

Existing dwellings have the potential to be transformed by modules. Remove the roof and add modules to turn from single storey to two storey. Or demolish part of the house and replace with a two storey modular segment. However have to have true modular. The buyer has to know what the module costs, and that it will be delivered and installed in a short time frame and that there are no additional hidden costs.

From another perspective houses are declared to be our most expensive purchase, but it seems the suppliers go out of their way to make them so. A module the size of a car, is mostly empty space, and should cost less in materials and labour than that of the car. The most involved spaces are kitchens, bathrooms and laundries: these need plumbing, wiring, various fixtures and cupboards, and surfaces tiling. But again they are not as involved as the components of a mechanised vehicle.  So where is the cost and what is the production delay?

Production

So from the above videos it seems one team can produce and empty shell in 1 day. So given 250 productive days in a year that one team has the potential to build 250 units per year. However, with CFS framing two frames can be built each day, increasing to 500 units per year, but these lack cladding. With the wooden shed, there were 6 workers, The basic tasks comprise of:

1. Floor Assembly
2. Wall Assembly
3. External Wall Cladding
4. Windows and Doors
5. Roof Assembly
6. External Roof Cladding

To a certain extent only need one person per task, but assuming need an assistant with each task, then need two people per task. If all tasks in parallel that would increase to a need for 12 people, though the assistants could move between tasks as needed if not required all the time. Staying with 6 people, split into 3 teams of 2, and adjust tasks into:

1. Floor Assembly
2. Wall Assembly
3. Roof Assembly

In this scenario, doors and windows, and wall cladding are part of the wall assembly, whilst roof cladding is part of the roof assembly.

Assume floor and roof stages of equal length, and a 480 minute day. Whilst wall stage longer than either floor and roof stage. So total time T=x + 2x + x = 4x, therefore x=T/4=480/4=120 minutes. So have stages:

1. Floor Assembly : 120 min.
2. Wall Assembly : 2*120 = 240 min.
3. Roof Assembly: 120 min.

So now can conclude can build 4 floor frames each day, and 4 roof assemblies each day, but only 2 lots of wall framing. So in a year, 2 workers have the potential to complete 1000 floor frames. If modular then the floor frames have a fixed width, but possibly varying lengths. So assume 4 frames are each 12m long, then they build 4*12=48m of framing in 480 minutes, or at the rate of 0.1m/minute or 10 minutes/metre.

So if work is staged in such manner then have a problem, as roofer cannot install roof until walls are built, and wall framer cannot install walls until floor built. However framing can be assembled at the same time as the floor is assembled. Assume this accounts for half the wall assembly time. So when the floor assembly is finished the walls are ready to install. To do this then require at least 4 work stations not 3, in a factory, if moving around sites then a different matter. If moving between sites then the floor framers just move to the next empty site, but if in a factory with a fixed number of work stations then they will get held up, as they will move between available work stations faster than buildings are complete. Or if work is moving between work stations, then it will not flow, as the work will back up at the slow station.

So whilst making wall framing in parallel with floor assembly saves some time, and drops total time from 480 to 360 minutes, this is not sustained unless have another two workers: 2 workers for wall frame assembly and 2 workers for installation. Otherwise whilst walls are being installed another floor frame is built, but there are no walls ready to install on it. whilst the wall frames are being assembled another floor is competed. So with 6 workers, it takes 360 minutes to complete the first building, and then 240 minutes there after. With 8 workers it takes 360 minutes to complete the first building and then 120 minutes there after. With 6 workers there is a delay waiting for wall assembly to install. So once the production line is flowing then producing buildings at the rate of 1 every 120 minutes. So in a 480 minute day that is 4 buildings per day and 1000 buildings per year. From 8 people in a factory compared to 6 people on site. Depending on building and equipment used then one person can make  the delivery. Though chances are one person cannot make more than 1 delivery per day depending on how far the destination is located.

So this is just a simple exercise without any real process data. However, it indicates that can change a system from 250/6= 41.6 buildings/worker each year to 1000/8= 125 buildings/worker each year. That is approximately 3 times the output, for 8/6=1.33 increase in workers. Or from another perspective we do not need more workers. Taking note that I deliberately made the intermediate task of wall framing take longer to create a delay. We can already assemble prefabricated wall frames and roof trusses on site so already have people producing these things. The problem is with the on site tasks, and many of these tasks can be taken off site.

Also workers don't have to get bored in a factory, they can follow an instance of the product down the production line, and then loop around again. This way they are involved in construction of an entire building, and using a multitude of skills.

Consider a larger building which takes longer. So this time all 3 stages take 1 day, and making use of prefabricated components. Each stage involves 2 people, so total of 6 people, and takes 3 days to produce one dwelling. But after completion of each stage the workers can move on to next project, this is what they typically do anyway, or at least some of the subcontractors do. So one team can make 250 floor frames each year. During the first year, the wall assembly and roof assembly people are behind, they finish the 250th building in the second year of operation. The question is can these 6 people produce more simply by changing their schedule? Based on the data the answer is likely no.

Also note that whilst it takes 3 days to produce a building, production is not limited to 250/3=83 buildings each year. Each building takes 3 days, but buildings are completed at the rate of 1 every day, after the first building is complete. In the time it takes to construct a building, 3 floor assemblies are produced, and 2 wall assemblies are installed, and 1 roof assembly installed. So on day 4, the second roof assembly is installed and a second building is complete, on day 5 the third roof assembly is installed and the third building is complete, and so on, a building is completed each day.

In the first case had 6 people on site building one building, and the assumption was these people are not fully occupied with construction. Therefore the people were either waiting or occupied with some task not directly related to the building shell. So the assessment was making them fully occupied on the building shell. If they were waiting to work then the waiting has been eliminated. If working on other tasks then those still need doing, but that also means time for the building shell is less than assumed, and therefore the shell can be built faster.

Tiny Houses

To improve on the 3 days the task needs breaking down into smaller stages. It is noted that tiny house builders are indicating 4 to 12 weeks to supply. Now that seems acceptable if they have a list of existing orders to complete. So as a time frame its acceptable, as a production duration it does not seem acceptable.

So a closer look at production stages:

1. Floor Assembly
2. Wall Assembly
3. Roof Assembly
4. External Roof Cladding
5. Electrical Systems (1st)
6. Plumbing Systems (1st)
7. External Wall Cladding
8. Internal Wall Linings
9. Internal Ceiling Linings
10. Electrical Systems (2nd)
11. Plumbing Systems (2nd)
12. Windows and Doors
13. Internal Decoration
14. Bathroom Fitout
15. Kitchen Fitout
16. Quality Checks

As only considering a small building, and there is no concrete to wait 28 days to cure, expect most of the tasks can be completed in less than 1 day. So at a maximum 16 days. With 16 buildings on the production line, then buildings would be completed at the rate of one every day. But already indicated that 5 of those activities can be completed in one day, leaving 16-5=11 activities, so total of 12 days of production. Still need 16 buildings on the line, and still only flow at the rate of one each day.

The problem, they only start building when get an order. So it takes 12 work days, which is short of 3 calendar weeks. Or if stay with 16 work days, then short of 4 calendar weeks. So unlike cars the buildings are not being produced to fill distributors show rooms. Or put another way there is no supermarket like supply of dwellings.

Supermarket for Houses

As the most expensive purchase in a persons live, it may not seem appropriate to buy a house from a supermarket. But it appears to be the most expensive item, because go out of the way to make it the most expensive item. A dwelling largely comprises the following spaces:

1. Kitchen
2. Bathroom
3. Laundry
4. Dining
5. Lounge
6. Living/Family room
7. Bedroom

The kitchen, bathroom and laundry are the main functional rooms, and not used all the time, so they can be shared amongst others. For many the main purpose of a dwelling is somewhere to sleep, which gives rise to the Japanese capsule hotels. Also in a sole occupancy dwelling, the lounge, dining, and living and sleeping areas are all shared space. In a caravan space is shared between the various functions depending on time of day. For some people all these extra rooms are a waste of space, and a waste of time keeping clean.

Providing people with private space to sleep and/or relax is easier than providing with an entire house. Providing shared facilities for washing and meals is also easier than providing to everyone. It is also possible to build a house such that the wet areas are a separate module to the rest of the house. The wet area module can remain on site, attached to services, whilst the rest of the building relocates as needed by the occupants. So a residential site has to be developed but it doesn't have to have an entire house on it.

If consider modular housing then most modules would be empty space, with wet area modules the only ones which require any significant fitout. So a wet area module may take 16 days, but the empty modules expect to take less, as the tasks would be reduced to:

1. Floor Assembly
2. Wall Assembly
3. Roof Assembly
4. External Roof Cladding
5. Electrical Systems (1st)
6. External Wall Cladding
7. Internal Wall Linings
8. Internal Ceiling Linings
9. Electrical Systems (2nd)
10. Windows and Doors
11. Internal Decoration
12. Quality Checks

So 12 days, or 12-5=7, so total of 8 days assuming shell only takes 1 day. So could produce shells which are taken aside and then modified into wet area modules, except that plumbing needs to be provided in the walls, floor and/or ceiling space, therefore really needs doing before a module is fully lined. So wet area modules likely better produced completely on separate production lines. If make fully modular then, kitchens, bathrooms and laundries made as separate modules, which can be assembled into larger more complete wet area modules.

So expect modules produced in large quantities and available mail order or from local show room. Is there a market? Not sure about current statistics but state of housing report, indicated that at any time some 5% of houses get a bathroom, kitchen renovation, or some outdoor renovation such as a deck or verandah, each year. Also some 5% of houses have inadequate number of bedrooms for the occupants.

So one issue there, is that real estate agents need to encourage people to move into more appropriate neighbourhoods and dwellings. Beyond that then have to modify the building that stuck with. Though with a lot of effort and reinforcement could lift the building off the site and relocate. The problem is a place to relocate to, and preparing the site. A true relocatable or transportable building can be located to a warehouse or a storage yard, the site doesn't require preparation if building in storage.

So consider a bathroom extension. Want to add an extra metre to the bathroom and refit it out. If it is close to boundary fence, then probably cannot do it in the first place, as there are fire safety issues and access to space between building and boundary fence. So expect have space for the extension and maintaining the boundary offset.

A module can take two forms. One option is it merely provides the 1m extension, the other is it replaces the entire bathroom. To replace the entire bathroom the module has to slide in horizontally or vertically. To slide in horizontally there needs to be adequate space besides the bathroom location. To slide in vertically the roof needs removing. So option 1 which only provides the extension is likely the simpler option. Also if module has its own floor then it will likely raise the floor level of the bathroom above the rest of the house, when really want wet areas below. So if have a concrete slab, will likely have to cut the slab out, which may have to do to modify the plumbing in any case. Alternatively the module needs to be constructed without a floor, and possibly without walls: in which case what is the module?

Whilst don't want to waste materials and wall thickness, double stud walls are typically installed to provide acoustic and vibrational isolation between rooms. So a module could just be a floor with fixtures on it and it uses existing walls, or it could come with its own walls. A floor could be removed and just have walls to the module if it is otherwise braced during transport and handling, with fixtures hanging from the walls.

If the house has timber floors then removing and replacing with module floor would be easier. Also if house already light weight construction or transportable than extending with modules is easier.

So the issue is to look at how house needs to be extended and modified and compare to modification using stock standard off-the-shelf modules. So masonry houses are the main issue, however most houses are brick veneer, and the existing bricks can be removed and cut into tiles and then replaced covering a greater area. The timber frame is also relatively easy to modify within certain limits.

So modules are useful if we can see how to use them and achieve our objectives.

This article is jumping around topics so will leave as is for now, and discuss other issues in separate posts.

---

Disclaimer:

I don't work for nor do I get sponsored by any suppliers I may mention. Nor is my mentioning a supplier a recommendation of their goods and services. Use the information at own risk.

---

Related Posts:

...

---

Revisions:

[31/12/2024] : Original

[1/1/2025] : Added Headings, some rewrite, and some extension