Platform approach (P-DfMA)

How does Platform approach (P-DfMA) work, step by step?

1. 1

   Step 1: Establish the portfolio case

   The client examines its forward programme and asks whether there is enough repeated demand to justify a platform. The relevant questions are the number of buildings, the period over which they will be built, how similar their requirements are and whether the client can commit to the platform across them. If the answer is a handful of buildings over an uncertain timescale, the platform approach is the wrong tool and component DfMA is the right one. This decision belongs to the client, not to any individual project team.

2. 2

   Step 2: Define the platform scope

   The client and its advisers decide what the platform covers. It might be the structural frame and its connections only, or it might extend through floors, walls, facade systems, services modules and fit-out. A narrower platform is easier to establish and control; a broader one captures more benefit. What matters is that the boundary is explicit, so that everybody knows which decisions belong to the platform and which remain with each project.

3. 3

   Step 3: Design the components and, above all, the interfaces

   Each component in the kit is designed once, properly, with manufacture and assembly in mind. The interfaces between components are then designed as first-class deliverables in their own right: what connects to what, how, with what tolerance and what adjustment. Stable interfaces are what make a platform a platform. They allow a component to be improved, re-sourced or replaced later without redesigning everything around it, and they are the reason a platform survives its first few projects.

4. 4

   Step 4: Prototype, test and validate the kit

   Components and connections are prototyped and tested against the performance the platform claims - structural, fire, acoustic, thermal, weathertightness and durability as applicable. This is done once, for the platform, rather than repeatedly for each building. A full assembly or demonstrator is often built to prove that the components go together as intended and that the assembly sequence works with real people and real tools. Problems found in a demonstrator are cheap; the same problems found on the third building are not.

5. 5

   Step 5: Build the supply chain and the manufacturing capacity

   The client establishes how the platform will be made and by whom, with enough capacity for the forward programme and enough resilience that a single manufacturer's failure does not stop it. Because the components are defined by the platform rather than by any one supplier, more than one manufacturer can generally be qualified to make them, which is one of the approach's structural advantages over tying a programme to a single system supplier.

6. 6

   Step 6: Configure individual buildings from the kit

   Each project is then designed by configuring platform components to the site and the brief, rather than by designing components. The design team's effort moves to arrangement, form, external envelope and the things that genuinely differ between buildings. Where the brief cannot be met by the kit, the exception is raised formally to whoever governs the platform rather than solved quietly on the project - which is the point at which most platforms are either preserved or lost.

7. 7

   Step 7: Govern change to the platform

   A named owner controls the platform. Proposed changes are assessed for their effect across the whole programme, not just the project that raised them, and accepted changes are versioned and issued so that every project knows which version it is building. Without this governance, each project makes small local improvements, the components diverge, the interfaces drift, and within a few buildings there is no platform left - only a set of similar-looking one-offs.

8. 8

   Step 8: Measure, feed back and improve

   Performance, cost, programme and defect data are collected from each completed building and fed back into the platform. This is the compounding benefit and the reason the approach is worth the effort: the fifth building genuinely should be better, cheaper and faster than the first, because the same components have been made and assembled repeatedly and what was learned each time has been captured in the kit rather than in one team's memory.