Construction keeps starting over: the $1.6T elephant in the room

This article is part of Brighteye’s Heavy Industry Series, a three-part exploration of how startups are transforming the physical economy. 

In this first piece, part two, we look at construction's often-overlooked workforce layer. Part one dove into manufacturing skills gap and part three will look at industrial engineering.

-----

Each construction project brings new people and new expectations, but when knowledge doesn’t flow and compliance gets reinterpreted on every site, the sector repeats its most costly mistakes. A new layer of tools is helping teams preserve both insight and integrity.

Introduction: The Hidden Cost of Starting From Scratch

Despite billions poured into construction tech over the past decade, many projects still end over budget and behind schedule. The reason for this being that industry forgets what it already learned in previous projects.

According to PlanRadar, as much as ~30% of construction work is rework, costing up to €1.6 trillion annually worldwide. KPMG found that 70% of construction executives see poor data management and lack of standardised processes as major barriers to performance. Meanwhile, the UK’s Construction Quality Improvement Collaborative (CQIC) estimates that error costs the UK construction industry alone £21 billion a year - much of it preventable.The root issue for this isn’t necessarily a lack of intelligence, but the fragmentation and lack of sharing of it.

Construction operates in cycles: each project is temporary, teams disband when the work is done, and institutional memory rarely carries over. That’s a stark contrast to manufacturing, where process continuity is paramount. In construction, each project starts again from scratch, new site, new sub-contractors and new tools, resulting in repeating errors.

At the same time, compliance requirements are becoming more demanding, from fire safety and emissions reporting to digital documentation mandates and material traceability. KPMG’s Construction in 2030 report highlights escalating regulatory pressure and data fragmentation as key drivers of delays, reinforcing just how vulnerable field teams are when guidance and oversight don’t scale from site to site (KPMG, 2024). Interpretation and application of these standards is incredibly complex, and limited coordination between design and build phases introduces new failure points. By the time regulatory errors are identified on-site, they’re costly to fix. Tools like AI BOB and Genia are stepping in upstream, helping architects and engineers proactively identify regulatory risks before construction begins. This shift from reactive QA to proactive design validation is reducing compliance friction across the build lifecycle.

The Project Reset Problem

General contractors and subcontractors typically move on after completion of projects, taking with them not just headcount but institutional memory. Documentation tends to be inconsistent, with workflows rarely standardised, and lessons learned are seldom structured in a way that benefits future projects. This forces teams to solve the same problems repeatedly.

What’s more, workforce access and continuity are in flux. Skilled labour shortages persist across Europe, and attracting new talent into a sector with outdated workflows and limited progression opportunities remains a challenge. Without mechanisms for knowledge sharing, collaboration, or skills development across sites, performance bottlenecks remain in place.

A New Generation of Tools for a Persistent Problem

Startups are now emerging to solve construction’s project-based amnesia. Their tools don’t just digitise existing processes, they enable:

• Real-time knowledge capture

• Workflow standardisation

• Regulatory foresight

• Field-to-design communication

• Scalable team training

We have flagged some of the companies operating in this space below (this is NOT exhaustive!):

Potential opportunities for startups:

With this graphic in mind, we sort to outline 5 areas and angles of opportunity that startups could pursue:

• Cross-project knowledge platforms: Nobody has yet “won” the horizontal layer that retains and reuses learnings across projects (like a GitHub for construction).

• Compliance-as-a-service: Dynamic regulation engines that continuously update requirements across regions, integrated into BIM/CAD. We are starting to see solutions like AI Bob emerge within this space, but are yet to see a winner.

• Insurance + warranty integration: linked to this, startups embedding QA/compliance data into insurance underwriting could open new revenue models.

• Labour enablement marketplaces: Many platforms exist (PowerUs, Fixed), but the “LinkedIn for trades” is still open territory. There is potential for vertical SaaS + fintech models (certification, payroll, insurance). This category comes with its own challenges for scaling e.g. localisation of roles, but we still see opportunities to improve efficiencies within the labour matching and support space.

• AI copilots for site managers: Tools that surface real-time risks, compliance gaps, or productivity blockers by aggregating multimodal data (photos, schedules, IoT, comms).

Why This Matters Now

• Regulatory pressure is rising, with stricter safety, emissions, and traceability mandates.

• Cost of error is massive, and largely preventable through earlier detection and structured knowledge flows.

• Labour shortages continue and are expected to only increase, meaning better tools are required to train and support the existing workforce.

• ESG requirements are climbing, and construction plays a central role in national emissions targets.

The Value of Preserved Knowledge

Construction’s biggest liability isn’t just inefficiency, it’s repetition. Without knowledge continuity or compliance foresight, projects remain isolated and quality varies.

But the tide is shifting. Startups are embedding intelligence into workflows, enabling field teams to build on past experience rather than discard it. From smart QA systems and embedded learning to AI-powered code checks, a new construction tech stack is emerging - one that’s finally designed around the people doing the work.

This is Part 2 in Brighteye’s Heavy Industries Series.

• Part 1: Manufacturing – Factories Don’t Run on Machines Alone

• Part 2: Construction – Solving the Sector’s Memory and Compliance Problem