Why good projects start with good schedules — Before BIM, AI and digital twins

Before any project invests in Artificial Intelligence (AI), digital twins or the latest BIM tools, there's a simpler question worth asking: is the underlying schedule any good? Because no amount of sophisticated technology can fix a programme that lacks solid logic, realistic sequencing or credible assumptions. Technology can enhance a good programme, but it cannot compensate for a poor one.

I work in construction planning, and this is a pattern I keep coming back to. The industry is excited about AI, digital twins and 4D — rightly so. But amid that excitement, the quality of the schedule itself often gets overlooked.

Technology amplifies quality — it doesn't create it

A programme is more than a contractual requirement or a reporting document; it's the strategy for delivering a project safely, efficiently and predictably. When done well, it answers a few fundamental questions: what needs to happen, in what sequence, what depends on what, which activities actually drive completion, and where the risks lie.

Those questions remain the same whether you're looking at Primavera P6, a spreadsheet, or a digital twin platform. Technology makes the information more visible — it doesn't create sound planning where none exists. A visually impressive 4D simulation built on poor logic is just poor planning shown in three dimensions rather than two.

The hidden cost of poor scheduling

Many problems blamed on "unforeseen circumstances" trace back to poor planning. I see the same issues come up repeatedly:

Activities with little or no logical relationship to each other
Constraints used in place of genuine sequencing
Durations based on optimism rather than evidence
Missing interfaces between contractors and disciplines
Commissioning, testing or temporary works left out of the model entirely
Schedules built to hit a reporting deadline rather than reflect reality

At first, none of this is obvious. Progress looks fine, dashboards stay green, reports keep circulating. It's only later — as delivery progresses — that the critical path becomes unreliable, delay analysis gets harder and teams start making decisions based on information they no longer quite trust. The problem is rarely the software, but rather the schedule underneath it.

Garbage in, garbage out

That phrase has never felt more relevant in project delivery. Every digital output — earned value reports, forecast dates, dashboards, digital twins — depends entirely on the integrity of the programme feeding it.

A digital twin built on bad logic gives misleading forecasts. AI analysing an incomplete schedule produces unreliable insights. A 4D simulation connected to unrealistic sequencing will confidently show you the wrong construction sequence. Technology speeds up decision-making — but only when what's underneath it can be trusted.

Planning is becoming more visible — and that's a responsibility

Planning used to sit largely behind the scenes. Now planners are expected to support decisions across the whole project — commercial strategy, logistics, design coordination, risk, stakeholder communication. Visual planning through 4D has made schedules accessible to people who'd never have interpreted a traditional Gantt chart.

That visibility is a good thing, but it raises the stakes. If teams are making major calls based on what a programme shows them, the person building that programme needs to be confident it's built on sound logic.

Planning is still a human skill

AI can spot trends, automate repetitive tasks and process large amounts of data quickly, but what it cannot do is replace professional judgement. An experienced planner understands why one activity has to come before another, recognises a constructability issue on sight and knows how design maturity affects delivery. That comes from experience and collaboration — not automation. Technology should support that thinking, not stand in for it.

Digital maturity starts with the basics

Organisations often measure digital maturity by how sophisticated their software is. A more useful measure might be the quality of the information going into it in the first place.

Before asking whether a project is ready for AI or a digital twin, it's worth asking simpler things: is the logic actually complete? Does the critical path reflect how the project will really be delivered? Are the assumptions behind it clear and regularly checked? If the answer is no, more technology just means bad information travels faster.

Looking ahead

AI, BIM, 4D and digital twins are genuinely significant opportunities, and digital transformation in this industry isn't going away. But these tools are force multipliers — not substitutes for good planning. Clear thinking, realistic sequencing and solid logic have always been what projects depend on, and that's not going to change. If anything, as our digital capabilities grow, getting the schedule right at the start matters even more.

Before the intelligent models and connected data environments, there's something far less glamorous that has to come first: a programme that actually deserves to be trusted.