Bridging the Analytics Gap: What We're Presenting at AACE

On March 26, 2026, I'll be presenting at the AACE International Northeast Section Symposium. The presentation is titled "Bridging the Analytics Gap: Progressive Schedule Analysis in the Age of AI," and it addresses something that's been on my mind for the better part of 35 years in CPM scheduling.

The Analytics Gap

The construction scheduling profession has no shortage of standards. PMI gives us the Practice Standard for WBS, the Practice Standard for Scheduling, and the PMBOK Guide. AACE gives us RP 29R-03 for forensic schedule analysis, RP 49R-06 for time impact analysis, RP 53R-06 for schedule development, and RP 92R-17 for schedule planning and development. The DCMA 14-Point Assessment provides a widely adopted health check framework.

The standards are comprehensive. The challenge is applying them.

A thorough schedule review that systematically applies these standards across a contractor's monthly update — WBS quality, DCMA compliance, schedule changes, float path analysis, milestone variance, and a consolidated narrative — typically consumes a full work week. Most of that time isn't spent on professional analysis. It's spent on data extraction, compilation, cross-referencing, and formatting.

The result is predictable: reviewers apply the standards they have time for, skip the ones they don't, and the quality of the deliverable depends heavily on who happens to be doing the work that month. The analytical depth that these standards were designed to enable gets compressed into whatever fits the deadline.

That's the gap. Not a gap in standards — a gap in the practical application of those standards to real schedule review workflows.

Progressive Analysis: Foundation Before Substance

The presentation introduces a framework we've been developing: progressive schedule analysis. The core principle is straightforward — structural validation must precede substantive analysis.

Before you analyze what changed between updates, you need to know whether the WBS is structurally sound. Before you trace the longest path, you need to know whether the schedule passes its DCMA health checks. Before you write findings about milestone variance, you need to know whether the changes driving that variance are logic modifications, scope additions, or constraint manipulations.

Each analytical step builds on the findings of the previous one. A finding in the schedule comparison step — "contractor added 14 activities to the mechanical branch" — becomes a question in the float path analysis step: "do any of those 14 activities appear on the longest path?" And the answer to that question becomes context in the milestone variance step: "three of those additions now sit on the longest path to Milestone M-100, extending it by 12 working days."

That's cumulative intelligence. It's the difference between a report that presents seven independent analyses and a report that tells a coherent story about what's happening to the project.

Where AI Fits — and Where It Doesn't

The presentation addresses AI directly, because it's impossible to have a conversation about schedule analysis in 2026 without it. But the framing matters.

AI is effective at translating structured analytical findings into professional narrative language. When the data shows that a contractor's logic density dropped from 94% to 87% between updates, and the schedule comparison identifies 23 deleted relationships, a system grounded in that data can generate narrative language that a scheduler would be comfortable putting their name on. It can reference the relevant AACE recommended practice. It can flag the severity. It can suggest what the reviewer should investigate next.

What AI cannot do is replace the professional judgment that makes those findings meaningful. It can't tell you whether a contractor's explanation for a schedule change is credible. It can't assess whether a recovery schedule is realistic based on the trade labor available in that market. It can't sit across the table from a project owner and explain why a 12-day longest path extension matters to their certificate of occupancy date.

The schedulers who thrive will be those who understand both the standards and the tools — using technology to augment their expertise, not replace their judgment.

Five Principles

The presentation closes with five principles for responsible technology use in schedule analysis. Without giving away the full content, they center on transparency, professional accountability, standards grounding, epistemic honesty about what can and cannot be determined from schedule data alone, and the non-negotiable role of human expertise in the review process.

These aren't aspirational statements. They're operational principles that govern how progressive analysis systems should work — and they apply whether or not AI is part of the equation.

If You're Attending

The symposium is on March 26. The presentation provides 1 Professional Development Hour (PDH). If you're in the AACE Northeast Section region, I'd welcome the chance to continue the conversation in person.

For those who can't attend: the themes we're presenting on — progressive analysis, cumulative intelligence, the analytics gap in applying established standards — are topics we'll continue exploring here and on LinkedIn.

The question isn't whether technology will change our profession. It's who will shape how that change happens — and whether schedulers will lead the conversation or follow it.

---

Michael Kehoe is the founder of IP System 3 LLC and developer of IPSYS Analytics, a patent-pending schedule review intelligence platform. He has 35+ years of CPM scheduling experience, formerly served as a Primavera P6 reseller and certified instructor, and has taught scheduling at Drexel University, Rutgers University, and Villanova University.

*Learn more at ipsysanalytics.com*