Defining *Reference Design* \\ Jaimie Johnston MBE
Every major client with a national or global rollout faces the same problem: procuring a bespoke design for every site adds cost, time, and variation that nobody needs. In this 'In Short' edition of the Bryden Wood Podcast, Head of Global Systems Jaimie Johnston MBE explains Reference Design – the methodology Bryden Wood developed to solve it. Read Jaimie's key takeaway article below, or click play to listen.
What is a Reference Design, and why use one?:
Many major clients have national or global rollouts that are happening at a rapid pace. Often, these clients have to procure a bespoke design for every site, introducing unnecessary design cycles and complexity that adds cost and time to construction projects and ongoing operations.
Yet, from our experience, these clients have a very well-informed view on how their assets need to perform and the layouts and designs that deliver their operational outcomes. This knowledge and understanding lends itself to a repeatable solution that can comprise c.80% of the asset.
We call this a ‘Reference Design’: a highly optimised, site agnostic ‘core design’ for a portfolio client. Typically, a Reference Design will contain a number of repeated elements at varying scales (from key equipment to whole assets) that can be configured in a number of different ways (with more or less flexibility in how these are arranged depending on the asset type).
For some assets, the Reference Design can become the fixed design. The house blocks we designed for the Ministry of Justice prisons rollout are a good example; highly standardised, they are identical on every site, only the number and orientation changes and is dictated by the prison population and aspect/prospect of the site. The UK prisons rollout is reporting dramatic reductions in delivery schedule with much improved certainty on just the second project to use this approach.
However, in other cases, where assets need to respond to site constraints and/or business needs, then the ability to adapt a Reference Design is critical to its usefulness. We have seen clients who already have Reference Designs (also called template designs). However, with no rules for how to adapt this Reference Design to local sites, the delivery teams can end up unravelling it and are back with new designs.
Many, if not most clients, need designs that can respond to unique sites and needs e.g. for a healthcare facility the design needs to be flexible to accommodate specific clinical specialisms and the demographics of a particular region.
So how can a Reference Design both standardise a design, yet leave enough flexibility to adapt it to any given brief?
Designing flexible Reference Designs
In Reference Design, we create a core design of the most common of our clients’ facility type using a process of rationalisation, optimisation and standardisation. At the same time, we attach a planned strategy for increasing or decreasing scale and content and how to flex to suit local conditions e.g. to scale up or down to hit a particular brief
rearrange or reconfigure to suit site constraints and layout
choose between a ‘seismic’ or ‘non-seismic’ version
choose between key suppliers for major equipment or specific materials required by fire codes etc.
We achieve this level of flexibility by thinking about the asset and the design in terms of ‘Chips’. To create Chips, we break your assets, and the processes within them, down into the smallest, meaningful sets of interacting or interdependent components. Chips are created at all levels, from
Simplest level e.g. single components that can be optimised for material efficiency
An Assembly e.g. MEP cassettes, which efficiently combine key components and processes
Individual spaces e.g. a patient room, which has certain needs, regulations
Group of spaces e.g. complex departments such as operating theatres, where all aspects of the room are dependent on each other to work properly
Whole asset e.g. when a Reference Design can be repeated as a fixed design c.f. MoJ houseblocks
The image below is an example.
Figure 1: A healthcare example of the different levels at which Chips are created
The depth and breadth of value achieved is greater, the bigger and more complex the Chip. The more co-dependent elements involved, the more variables and complexity the Chip has to resolve and the wider the implications of the ways in which Chips will optimise the design. So while an inpatient room is optimised for patient comfort and the facilitation of day to day care activities, an operating theatre has more complex variables to incorporate: medical procedures, heightened levels of infection control plus the associated surgical staff, flows of clean and dirty consumables, carefully controlled flows of air, extensive use of medical equipment and data etc.
The optimal design, which we call the Reference Design, can then be largely created by combining these Chips in the most efficient way. This Reference Design can then be built again and again as a fixed product, by clients who have control over site conditions, as per the MoJ Houseblocks.
But for those who need flexibility, the Reference Design becomes far easier to configure flexibly into a physical asset when broken down into its ‘Chips’ (with more or less flexibility in how these are arranged depending on the asset type.) By digitally modelling the design against the given brief/constraints using Chips, we can quickly see how well the design will work, and what we can adapt to fit the wide range of needs and constraints involved. This ability to adapt the configuration of a design would be impossible in a design expressed as 100% single components.
The ‘local’ design is then focused on ground conditions, utilities, infrastructure, placemaking, flows of people etc.
Figure 2: Example configurations of healthcare Chips
We introduced this concept to the New Hospitals Programme, developing the initial ‘Hospital 2.0’ solution. While this has since evolved, the use of a Reference Design continues to be a central part of their strategy to deliver schemes across the UK.
In addition to the public sector health and custodial programmers, we are currently deploying Reference Design and Chip Thinking across the private sector, particularly in the data centre market and for logistics and fulfillment centers.
The benefits of using a Reference Design even with a traditional supply chain
Sometimes the lack of supply chain maturity necessitates a more traditional construction process. However, even using a completely traditional build, reference design has numerous benefits:
It facilitates the introduction of industrialised construction (including platform approaches or P-DfMA) and the use of manufactured elements in construction
Often owned by the client, it gives greater control over IP and incorporates lessons learned to improve across design cycles
Multiple use allows for more design refinement, amplifying the benefit of good design
It justifies a greater level of stakeholder engagement ensuring that designs are highly optimised in terms of layouts, space allocation, adjacencies, and functional flows
Designers can focus more of their efforts on solving the site and context specific challenges
It facilitates efficient operation and maintenance
Critically, Reference Design gives teams the ability to assess a site very quickly using a ‘test fit’ process:
The client brief drives the selection of the appropriate Chips required to build the asset
These can then be arranged on the site according to the engineers’ predefined rules e.g. proximity of the Chips to one another; which ones can/can’t be stacked, which ones have to be next to each other vs separated etc.
This allows an initial solution (or multiple solutions) to be developed, with pre-defined rules assuring compliance
These design solutions can be assessed and compared for optimisation of flows, simplicity of enabling works (cut and fill, extent of utilities works) etc.
This is not a complete design solution but can quickly generate a working feasibility model so clients can make highly informed decisions in days or weeks and at a fraction of the cost of a non ‘reference design’ approach
Clients can even take a virtual tour to aid their insight, feedback and input
The design and procurement stages that follow can then also be compressed through focusing time and effort only on the bespoke elements of the largely standardised design
This approach has been enormously powerful for our clients: the ability to rapidly assess the viability of sites has saved months of uncertainty and prevented the inefficiency of wasted work.
Configuring Reference Design
The rule and data-led thinking that defines Chips and how they can fit together in a design e.g. scale, adjacency, proximity etc. lends itself to the use of Configurators to automate the design process.
This isn’t a necessity – many clients have reported that just the process of defining and writing down their own rules processes is valuable, and the manual test fits are already far quicker than bespoke designs.
However, codifying these rules to drive an algorithmic or computational design process, results in a dramatic increase in the speed and quality of adapting Reference Design to a myriad of constraints and needs. One of our previous projects, Seismic, for example, was a design configurator app that tested the suitability of new sites for schools by configuring the potential options in 10 minutes, an apartment block was configured in half an hour, a motorway in an hour. At this speed, we can beta test many more options than could usually be considered – one of our client specific configurators will generate many thousands of completely compliant test fits in a matter of minutes, to select the solutions which best suit a specific site.
Industrialising Reference Design
Reference Design is of immediate benefit to the current design to construction market; even if built using a traditional process, there will be benefits brought by the optimal design (via optimisation of adjacencies and the standardisation inherent in Reference Design).
However the greater opportunities lie in its ability to unlock DfMA and industrialised construction; using Reference Design for a programmatic roll out gives Portfolio clients an opportunity to move up the adoption scale at a pace that allows the supply chain to move confidently with them. The first assets may be delivered quite traditionally, but sequential assets can be increasingly industrialised in their delivery:
Starting to engage key suppliers in the provision of standardised equipment to give certainty of pipeline to the supplier, and certainty of cost and availability to the client
Prefabricating sections or ‘lower order Chips’ (the standardisation process means the lower order Chips are often designed in a way that means they can be prefabricated e.g. an MEP cassette)
Culminating in the development or adoption of a platform solution to the reference design.
Jaimie Johnston MBE is Director and Head of Global Systems at Bryden Wood.
