Delivering an Automated Precision Manufacturing Facility for Off-Site Housing Products

Berkeley Modular is building an advanced manufacturing facility for the purpose of producing volumetric modular product which will be used to develop new residential accommodation. The intention of this article is to explore why and how they are doing this. Accordingly, we will focus on the following topics:

• Creating a digitally enabled agile manufacturing platform to support automated data workflow for an off-site housing product

• Capturing structured product data for off-site housing products for use in design, manufacturing, and through-life support

• Using computational logic to obtain high-level design automation to create data-rich models, BOM and BOP data, and machine code

• Using harnesses for regression testing to support development of rules for design automation of off-site housing product

The Backdrop

Innovation drives revolution, and history shows us the potential scale of impact across industrial and transport sectors. Albeit off-site construction has been around for a while, it has generally failed to produce the same sort of revolution, and there are lots of reasons for this. In the UK, The "Farmer Review of the UK Construction Labour Model," a report published in 2016, provided a reminder of these reasons and challenged the construction sector to recognize their failings and think about what sort of transformation was needed to bring about long-term revolution.

In point of fact, statistics reveal that construction is no more productive in the UK today than it was 10 years ago. Given that many players in the off-site sector have been around for decades, it is not unreasonable to conclude that the existing approaches to off-site are also part of the construction productivity problem. In the 2019 McKinsey Report, “Modular Construction: From Projects to Products,” the productivity conundrum is laid bare, as is the importance of putting off-site at the forefront of business operations to drive a productivity revolution.

Over the last one to two years in the UK, a new breed of players have entered the market, including Berkeley Modular, and these players are bringing a different emphasis to bear and that is why we could be at the dawn of an off-site revolution. This change of emphasis is based on the newer players being focused on off-site manufacturing as opposed to off-site construction, and there will be more details pertaining to this distinction later.

Berkeley Group has been around for some 45 years, and on average develops around 4,000 properties per annum. Berkeley Group faces the same challenges as the rest of the construction sector, as highlighted in the Farmer Review, so when the decision was taken to enter the off-site sector it concluded it needed to be at the forefront of the field. Whilst the decision to invest in its own manufacturing capacity is related to surety to be able to continue to build 4,000 properties per annum, as Berkeley Group has a significant pipeline of properties to build, it was also considered important to change the skills dynamic and be able to manufacture product where actual performance meets designed intent.

The first advanced manufacturing facility being developed in Northfleet, Kent has been conceived to produce 20-25% of Berkeley Group’s annual output at peak capacity. The volumetric modular product to be produced in the new facility will be deployed to create residential housing with a bias towards apartment living of 3-20 storeys.

The nature of the product development work that has been carried out over the last two years or so has focused on creating a product solution that does not constrain architectural freedom (i.e., the intent was never to devise a limited catalog of template product) but is designed to support manufacturing and assembly logic, and is mass customizable.

Most significantly though, through a focus on lean thinking and continuous improvement, the investment is intended to help drive productivity in a sector that has shown little evidence of fundamental change over the last 10-20 years. Consider, a volumetric module that is being produced to the maximum geometric dimensions that production process can accommodate might weigh 20 tonnes, and the conversion time from piece parts to tested product at end of line will eventually be 40 hours.

Berkeley Modular is a newly formed division within Berkeley Group and was created in April 2017. The volumetric product is a steel-framed based solution, and the new facility comprises 150,000 sq. ft. of production space, 15,500 sq. ft. of office space and 220,000 sq. ft. of yardage / hardstanding.

One question that is regularly asked is why the space afforded to yardage / hardstanding footprint is greater than the production space and the answer is straightforward: the logistics challenge in running a lean manufacturing business is a limiting factor and material feed to the process running at peak capacity and producing 12 modules per day implies a daily replenishment requirement for material of circa 200 tonnes. The process logic for the new facility is much more akin to manufacturing (e.g., automotive, aerospace, etc.) than construction.

Even at the most basic level, the manufacturing sector bears little resemblance to the construction sector. Significant differences exist between the two, typically manifest in terms of culture: operating philosophy; productivity; return on investment; employment and talent development rationale; and so forth. For some reason though, when ‘off-site’ is the prefix to manufacturing or construction, people often consider the resulting terms to mean the same thing. However, they do not, they imply very different things. This confusion regarding the terms off-site manufacturing and off-site construction suggests it is worth attempting to differentiate between the two.

Whilst the UK has a long history of off-site, much of the activity over decades has leaned towards off-site construction-based approaches. As has been evidenced in other sectors though, the radical transformation of productivity, the scope to drive improved, consistent quality, and the opportunity to quickly scale are only made possible by thinking differently. The key point is to recognize that an off-site manufacturing approach facilitates predictability and repeatability, and more readily affords scope to embrace digitization with an emphasis on Design for Manufacture and Assembly (DfMA) as opposed to just visualization.

By applying the right sort of thinking it is possible to envision a flexible off-site manufacturing methodology which can support the notion of mass customized product (i.e., non-template / non-platform solutions) with capacity for high conversion velocity (i.e., the elapsed time to convert raw materials to finished product). For Berkeley Group, these sorts of outcomes helped to provide the rationale for making the investment in capital equipment and developing a different sort of talent pool that might provide the foundation for a transformative industrialized logic.

For Berkeley Group, the decision to invest was a major milestone and rather than simply replicate traditional construction activity in a controlled environment it was decided that a bolder step was needed. Some 200 years ago, we experienced the industrial revolution which transformed the UK’s economic landscape. Then, some 100 years ago, we experienced the transport revolution which further transformed the UK’s economic landscape. Today we are witnessing the emergence of the off-site revolution which will transform the construction sector landscape and bring further economic uplift. The benefits to the construction sector are manifold:

• Leverage productivity

• Increase quality

• Increase predictability

• Eliminate structural fragmentation

• Address skills shortages

• Leverage collaboration

• Enhance sector reputation

• Embrace technology

• Create a sustainable future

Digitally Enabled Agile Manufacturing

As with other countries around the world, the UK faces a deep-rooted housing crisis and the UK government has been bringing increasing pressure to bear on developers to leverage output, and address the increasing lack of skilled resource available to the construction sector. In simple terms the various initiatives launched by the UK government have been focused on leveraging higher levels of better quality output for less input. So, the strategic appetite to embrace digital technologies has shifted over time, but recent events in the UK have brought fresh impetus to bear.

Of particular note over the last couple of years is the attention that has been brought to bear on capture of the golden thread of building information, in other words the digital record should not just reflect what was built but also how, when, who and why it was built. Faced with the problem of having to explain a complicated notion it is sometimes helpful to draw on an analogy. 

The conundrum of how to capture the golden thread of information through a design and build cycle to properly support effective operation and maintenance of a residential development could rightly be considered one such complicated notion. The idea of a river as the natural phenomenon that has parallels with the task of capturing the golden thread of information is useful.

In recent times, there have been advances in technology such as digital design software solutions and common data environments that can be used to help create a complete and accurate digital record for a residential development. These tools can be considered to represent additional artificial constructs that can help facilitate control of information flow, so in a sense they are also akin to the concept of channeling a river between watersheds.

But equally, there have been important changes in the way parties engaged on a project work together which are also yielding influence. Furthermore, we have seen the emergence of new standards and codes of practice associated with the likes of naming conventions, common language definition, data exchange, and building of information models, all of which also constitute artificial constructs which are intended to make the process of capturing the golden thread easier. This wealth of change in working practice is equivalent to adaptations in topography in the analogy.

Under normal circumstances, it is easy to comprehend how the fragmented nature of conventional construction approaches cause complexity in terms of information authoring and liability, which acts to thwart and frustrate the process of capturing the golden thread of information. This complexity might arise due to a lack of foresight regarding the need to capture and manage information from potential sources from the beginning, or from a lack of application of the artificial constructs required to control information flow over the design and build cycle. In contrast to the regenerative nature of a river ecosystem, it is often the case that the lack of application or even inappropriate artificial constructs can lead to the evolution of a form of extractive process which is overly linear with many disconnects and embedded wasteful logic reflecting an ineffective approach to capturing the golden thread.

Berkeley Modular is a business focused on the off-site manufacture, as opposed to off-site construction, of three-dimensional primary structural product (i.e., Category 1 in accordance with the UK government’s MMC definition framework). We have been afforded the luxury of time to conceive how we can apply lean thinking to information authoring, capturing and revision control, as well as to our manufacturing and assembly logic.

The Digitally Enabled Agile Manufacturing (DEAM) platform we set out to create has several component parts as set out below.

• Digital capture of information from source–We have deployed options from the coBuilder suite of software to configure the DEAM platform to facilitate a single source of truth for all assets to be incorporated in a project. These options represent the tributaries that allow information to be authored by manufacturers’, and subsequently filtered and fed to other components of the DEAM platform.

• Digital creation of design information–With the help of Majenta, we have deployed certain options from the Autodesk suite of software to configure the DEAM platform to facilitate a product lifecycle management tool wherein digital geometries and build logic definitions are automatically linked to asset information in a common data environment. These options represent the topography that allow Berkeley Modular to author design and build definition, and subsequently filter and feed to other components of the DEAM platform.

• Digital creation of manufacturing instruction–Working with DAS, we have deployed computational rule-based logic to obtain high levels of design automation to support the efficient creation of data-rich, fully federated digital models and related machine code. This logic represents watersheds that afford authoring of automated build definition by Berkeley Modular, which can be filtered and fed to other components of the DEAM platform.

• Digital management of supply chain, operations and finance activity–We have deployed certain options from the Oracle Fusion suite of software to configure the DEAM platform with an ERP environment which facilitates a single source of truth for all aspects of operational activity. This environment represents further topography that allows capture of all transactional information authored by Berkeley Modular, and subsequent filtering and feeding to other components of the DEAM platform.

• Digital instruction of manufacturing and assembly activity–We have deployed certain options from the Siemens suite of software to configure the DEAM platform with an MES environment which facilitates a single source of truth for organizing and communicating all facets of physical activity performed by Berkeley Modular. These options represent final elements of topography that afford capture of work instruction to both machine and human resource across all factory and site operations, which can be filtered and fed to other components of the DEAM platform.

The simple logic of the DEAM platform is that it will help reduce the time required to produce high quality design / technical output and convert to work instruction. Consider, at peak capacity, a conventional approach to satisfy the requirement to publish the required output to manufacture 2,500 unique volumetric modules per annum might imply a headcount of 50 people.

This is based on the notion of each fully federated, data rich model we produce might require five days of resource time and so with only 250 working days available per annum implies a level of resource which would prove challenging—in terms of recruiting and also in terms of accommodating. In the alternative, the principles behind the DEAM platform are intended to support the notion of each fully federated, data-rich model we produce might require one day which would yield a much more practical headcount requirement of 10 people.

In the class, we will further introduce ways via which Berkeley Modular has set out to resolve some of the challenges in generating design output and how to use digital technologies to leverage productivity and quality in order to feed downstream processes. Productivity is not just an issue for the manufacturing elements of a business, it applies across the entire business.

Neil Lee is a director of Design Automation Systems, a software consultancy that has specialized in rules-based automation of design and engineering for over 20 years. Prior to this, Neil did a combined software / engineering degree and was a director of a consultancy focused on finance, ERP and accounting software.

Graham Cleland is managing director of Berkeley Modular, a part of the Berkeley Group which is one of the UK ‘s leading residential developers. Before founding Berkeley Modular, he amassed a wealth of experience in the construction and engineering sectors in a career spanning some three decades.