Oil and gas · Consultancy

AI beat the human forecast on 15 of 17 milestones.

PaCE uses Nodes & Links AI to remove optimism bias from schedule forecasting, improving accuracy and enabling proactive risk management.

Book a demo

15 of 17

milestones forecast better by AI

37%

more accurate milestone forecasts

Up to 10x

faster decisions

About PaCE

Objective forecasts that complement human expertise

PaCE is a consultancy specializing in project controls across mining, oil and gas, petrochemicals and infrastructure, helping clients plan, forecast and manage complex projects.

This story covers PaCE’s work in oil and gas, where high-stakes, time-sensitive projects make accurate, actionable insight essential.

The project

Oil and gas projects, where downtime costs can reach millions per day
Human forecasts consistently underestimated the time and resources needed: optimism bias
PaCE needed objective forecasts to direct mitigation where it mattered most

The challenge

“Which milestones carry the highest risk of delay?”

Optimism bias: human forecasts kept underestimating the time and resources the work would need.

The solution

From analysis to a clear decision

The team used QSRA to get to the answer, with every result traceable to the schedule.

Monthly data collection: milestones, activity updates and forecast changes
Each month, PaCE gathers up-to-date project data, tracking milestone completions, activity updates and any changes to forecasts.
AI analysis: a monthly QSRA with Nodes & Links
The AI runs a quantitative schedule risk analysis to identify deviations between forecast and actual timelines, a monthly analysis that would be hard to sustain with traditional methods.
Deviation identification: compare AI and human forecasts
Milestones with the largest gap between the AI and human forecasts are flagged for investigation. Small deviations are deprioritized.
Expert mitigation: targeted risk prioritization
PaCE experts assess the flagged milestones and activities and develop targeted mitigation, so resources go where they are needed.
Assurance reporting: every month, for PaCE’s clients
AI-driven findings are compiled into monthly assurance reports, giving clients actionable guidance to manage risk and maintain project timelines.

The result

AI forecasts averaged 15 days of error per activity against 41 for human forecasts, and experts focused mitigation on the milestones with the widest gap.

15 of 17milestones forecast better by AI37%more accurate milestone forecastsUp to 10xfaster decisions

Average forecast error per activity.

Human forecast

41 days

AI forecast

15 days

“The ability to measure deviations between AI and human forecasts has transformed our risk management strategy.”

Byron van Belkum, Associate Director, PaCE

Show your working

Every answer, explained

An answer is only useful if it holds up. Nodes & Links records the reasoning and model outputs behind every answer, so the team can see each step, check the math and share a recommendation that’s traceable.