What is project controls?
Project controls is the discipline of keeping a project's time and cost outcomes within plan. It is distinct from project management: the PM decides and directs; project controls measures, forecasts, and warns. On a well-run job the two form a loop — controls tells management where the project is and where it is heading, management acts, controls measures the effect.
The discipline rests on five pillars:
- Scheduling — building and maintaining a network of activities with dependencies, calendars, and a critical path that actually reflects the delivery strategy.
- Progress measurement — capturing what was actually done, at the activity level, at a cadence fast enough to matter.
- Risk management — identifying what threatens the plan, quantifying the exposure, and tracking mitigation.
- Performance reporting — turning schedule and progress data into honest, current answers for owners, leadership, and clients.
- Change and escalation management — a governed path for issues, changes, and decisions, so problems move up rather than sideways.
Most organisations have all five on paper. The failure is in the plumbing: each pillar runs on its own tool and its own cadence, and the reconciliation between them is manual. That is why classical project controls is a lagging discipline — by the time variance is measured, reported, and discussed, the window for cheap correction has closed.
How Gantivity implements each pillar
Scheduling: AI generation, continuous critical path
Gantivity Studio generates the schedule from plain-language scope — WBS, dependencies, durations, calendars — and computes a Gantt-backed critical path. Because generation is cheap, every package and recovery plan gets a real network, not just the main contract programme. Details on the AI construction scheduling page.
Progress measurement: daily, structured, from the work face
The Daily Progress workflow captures notes, percent progress, and time per task, from the site engineer's phone, with full history. Measurement stops being a monthly reconstruction and becomes a daily byproduct of the work — which is the single change that makes every other pillar leading instead of lagging.
Risk: delay prediction from execution signals
Execution signals classify every task as on track, behind, or blocked, and the delay-prediction layer compares actual performance against planned rates to flag activities and milestones trending late — before the float is gone. Risk stops being a static register reviewed monthly and becomes a live overlay on the schedule itself.
Reporting: generated, not assembled
Status reports, risk summaries, weekly digests, and schedule-variance narratives are generated by AI from the live schedule on demand. ZAI, the in-channel copilot, answers ad-hoc questions — due this week, tasks with issues, channel summaries — in seconds. The controls team reviews and interprets; nobody spends two days a month formatting.
Change and escalation: governed by default
The issues register carries priority and severity, and configurable auto-escalation moves unanswered critical items up the chain automatically, with a log of who was notified and when. Project controls messaging (PCM) keeps structured nudges aligned to pending work, so the governance runs in the same place the conversation does.
