To create a built environment system that can continuously learn, retain knowledge and improve over time, construction must learn five key lessons. So argues Sam Stacey, director of A7C and former director of the Transforming Construction Challenge from 2018 to 2022, in his new book, Brunel’s Bees.
This is not a paywall. Registration allows us to enhance your experience across Construction Management and ensure we deliver you quality editorial content.
Registering also means you can manage your own CPDs, comments, newsletter sign-ups and privacy settings.
And yet, the most striking feature of this situation is that the solutions are not unknown. We already understand how to deliver dramatically better outcomes: lower energy demand, fewer defects, higher productivity and more reliable performance in use. The real challenge is not invention, but adoption.
Over the past decade, this has been tested in practice. Through the Transforming Construction Challenge (TCC), we demonstrated what becomes possible when the sector begins to operate more like a learning system – proving new approaches, building an ecosystem of innovators, and beginning to shift culture across industry and government.
The next phase is now underway. The Industrialising and Digitalising Construction Challenge (IDCC) moves beyond proof of concept towards scale, focusing on standardisation, digital marketplaces and the wider adoption of industrialised approaches. The IDCC represents the transition from experimentation to industrialisation.
But even this progression, important though it is, does not fully resolve the underlying issue. These programmes accelerate innovation and deployment, but the deeper challenge remains: how to create a built environment system that can continuously learn, retain knowledge and improve over time. That is the shift explored in my book Brunel’s Bees, and it is the foundation for the five lessons that follow.
Treat construction as a system, not a series of projects
The modern built environment is still largely organised around discrete projects. Each one assembles a temporary coalition of clients, designers, contractors and suppliers, only to disband once delivery is complete. The consequence is predictable. Knowledge is generated but rarely retained. Lessons are learned locally, then lost. Each new project starts again, repeating familiar problems. This is the central structural weakness of the sector.
Other parts of the economy have moved beyond this model. Manufacturing, for example, operates as a continuous system. Learning accumulates, and processes steadily improve. Performance rises over time. Construction must do the same.
This means shifting from a project-centric view to a system-level perspective where programmes, portfolios and institutions are designed to retain knowledge, align decisions and improve collectively. Without this shift, even the best innovations will remain isolated successes rather than drivers of systemic change.
Build the capacity to learn
The challenge is organisational, not technical. The built environment contains extraordinary expertise. Digital tools allow us to simulate complex systems, optimise designs and manage delivery with increasing precision. Yet the system often behaves as if it cannot see the whole. Responsibility is fragmented. Information is siloed. Decisions are optimised locally rather than systemically.
To understand what is missing, it is useful to look at systems that do learn. A beehive, for example, coordinates thousands of individuals through feedback, memory and shared signals. Over time, it adapts. It becomes more effective at sustaining itself. The built environment rarely achieves this form of collective intelligence – but it could. Learning systems require three things:
- feedback: data on performance in use;
- memory: mechanisms to retain and transfer knowledge; and
- coordination: alignment of decisions across actors.
Without these, improvement remains episodic. With them, it becomes continuous.
Integrate design, construction and operation
One of the most persistent divides in construction is between the phases of the asset lifecycle. Design is often treated as a standalone activity. Construction begins with limited continuity from what came before. Operation is disconnected from both, with little feedback into future projects. This fragmentation undermines performance.
In systems that function well, these phases are integrated. Design decisions are informed by operational data. Construction builds on prior learning and assets are treated not as endpoints, but as part of a continuous cycle of improvement.
Historically, this was not beyond reach. Nineteenth-century infrastructure programmes – often associated with figures such as Brunel – coordinated engineering, finance, materials and logistics at extraordinary scale. Knowledge accumulated as networks expanded, while techniques improved from project to project.
Today, we have far more advanced tools, yet often less coherence. Reintegration is therefore essential. Digital models, data platforms and long-term asset stewardship can enable continuity across the lifecycle. But this requires institutional alignment, not just technological capability.
Focus on outcomes, not outputs
The industry typically measures success in outputs – cost, time and scope – because these can be specified and verified at handover. But this obscures what ultimately matters: how places perform in use.
Brunel’s Bees reframes this through the lens of beauty. It is not a decorative extra, but a diagnostic signal of system coherence: evidence that alignment has been achieved. Attempts to mandate beauty through codes mistake an output for an outcome. Real quality emerges only when governance, skills, procurement and incentives are aligned around long-term value rather than short-term metrics.
Fragmented, risk-averse systems suppress the conditions that allow dignity, comfort and cultural meaning to appear. Conversely, when purpose and responsibility are aligned, these qualities emerge naturally, even under constraint.
This points to a deeper idea: flourishing – a condition in which performance, productivity, human experience and stewardship reinforce one another. The task is not to specify better outputs, but to design systems in which they reliably emerge.
Redistribute value to enable change
Finally, transformation requires a rethinking of how value is distributed across the system. This was addressed in the TCC’s Value Toolkit. Today, many of the benefits of improved performance – lower energy costs, reduced maintenance, better user outcomes – accrue over time and often to different actors than those who invest upfront.
This creates a structural disincentive to change. If the system is to evolve, value must be aligned with performance. This may involve:
- new commercial models that reward long-term outcomes;
- procurement approaches that prioritise whole-life value; and
- institutional arrangements that support continuity across programmes.
The system must be designed so that doing the right thing is also the rational thing.
From fragmentation to coherence
Taken together, these lessons point to a fundamental shift. The challenge facing construction is to become a system that learns. One that can retain knowledge, align decisions and improve over time. This matters far beyond the sector itself. The ability to adapt the built environment will increasingly shape national outcomes. Housing, infrastructure, energy systems and public services all depend on it.
An industry that learns can respond coherently to these challenges. One that does not will continue to struggle. The encouraging reality is that the path forward is already visible. We know what is possible. We understand the principles. There are examples, often at the edges of the system, where these ideas are already taking shape.
The task now is to embed them. Because the prize is a transformation in how society builds, adapts and flourishes.
