Our Analysis Methodology
How ScheduleLens assesses construction schedule quality — grounded in published industry standards.
Analysis Approach
ScheduleLens performs an automated assessment of construction schedule files against published industry standards. Each finding in the report traces to a specific rule in one or more recognised standards.
The analysis is deterministic — the same file always produces the same results.
The role of AI. Every finding, score, figure and chart in your report is computed by deterministic analysis code, not generated by an AI model. Each result traces to a specific rule in a published standard. In your report, AI is used in one place: a language model (Qwen 3.6 27B, run by Venice.ai) turns those computed findings into readable narrative prose, working only from the computed results. The report's Methodology section names the sections it wrote, and where prose and a table or chart differ, the table or chart governs. It is optional: switch it off for one upload, or for every upload under Account > Report preferences, and your report uses template prose with nothing sent to any AI system. Either way, the findings are identical. The model, what it receives and its limits are set out on our AI transparency page.
The tool identifies quality issues; it does not determine fault, entitlement, or cause of delay. Where a report shows a net EOT position, it is an arithmetic aggregation of responsibility classifications you assigned yourself on the delay register — the tool suggests (clearly labelled as preliminary) but never classifies for you, and events you have not classified are always shown as Unclassified, widening the stated range rather than being excluded.
The Quality Baseline
Ten categories of schedule quality assessment, each grounded in published standards. This is the layer every schedule checker covers; the integrity signals below are the layer most do not.
Logic Completeness
Verifies every activity is connected by predecessor and successor relationships.
Gaps in logic mean the schedule cannot properly calculate the critical path.
DCMA 14-Point, AACE RP 29R-03, PMI Practice Standard
Relationship Quality
Assesses the distribution of relationship types and identifies unusual patterns such as Start-to-Finish links or excessive lag.
Unusual relationships can mask real schedule behaviour and mislead stakeholders.
DCMA 14-Point, AACE RP 29R-03, PMI Practice Standard
Float Analysis
Examines total float distribution to identify activities at risk, near-critical paths, and potential missing logic.
Understanding float is essential for prioritising management attention on at-risk activities.
DCMA 14-Point, AACE RP 29R-03, PMBOK Guide
Constraint Audit
Identifies hard constraints that override logic-driven dates, potentially masking true schedule risk.
Constraints can hide the real critical path and give a false sense of schedule health.
DCMA 14-Point, AACE RP 29R-03, SCL Protocol, PMBOK Guide
Calendar Validation
Checks for calendar anomalies, cross-calendar relationships, and configuration issues that can distort working-day calculations.
Calendar errors silently corrupt duration and date calculations across the entire schedule.
AACE RP 29R-03, PMI Practice Standard, PMBOK Guide
Critical Path Integrity
Validates that the critical path stored in the schedule matches an independent recalculation. The disagreement is reported as a finding; the schedule's own values are never replaced.
Discrepancies indicate stale or unreliable data that cannot support delay analysis.
DCMA 14-Point, AACE RP 29R-03, PMBOK Guide
Duration Checks
Flags activities with zero remaining duration while in progress, and activities with unusually long planned durations.
Duration anomalies indicate data entry errors or activities that should be decomposed.
DCMA 14-Point, PMI Practice Standard
Progress Assessment
Identifies activities that are behind their planned finish or have started without recording progress updates.
Missing progress updates mean the schedule no longer reflects reality on site.
DCMA 14-Point, SCL Protocol, PMBOK Guide
Out-of-Sequence Detection
Flags activities that have progressed before their predecessors completed, which distorts the schedule's calculated dates.
Out-of-sequence progress undermines the reliability of forecast dates.
AACE RP 29R-03, PMI Practice Standard
Redundant Logic
Identifies relationships that have no scheduling effect because a longer path already exists.
Removing redundant logic simplifies the network for clearer analysis and reporting.
AACE RP 29R-03, PMI Practice Standard
Update-Integrity Signals
A monthly update can be made to read healthier than the works are: a settings change, a shaved duration, a forced date or a rewritten actual will do it. These checks look for exactly that. Each one states the mechanical evidence it rests on and leaves the judgement to you: the file can show that a figure changed, not why.
Scheduling-Mode Disclosure
Reads whether the file was calculated under Progress Override or Retained Logic, and flags a switch between updates.
The two modes can produce different critical paths and float on identical activity data. A reader comparing two updates needs to know the setting moved.
AACE RP 29R-03 §3.3.D, SCL Protocol §11, CIOB Guide §4.30
Concealed Slip
Recomputes a Progress Override schedule under Retained Logic and reports the difference in forecast completion.
Progress Override can hold a completion date static while the network behind it has slipped. The difference is what the setting is hiding.
AACE RP 29R-03 §3.3.D, SCL Protocol §11, CIOB Guide §4.30
Expected Finish Forcing
Counts incomplete activities carrying a per-activity Expected Finish date, and flags recorded progress that contradicts the forced date.
Expected Finish overrides the logic-driven forecast without registering as a constraint, so a constraint audit sees nothing.
AACE RP 29R-03 §2, SCL Protocol §1, CIOB Guide §4.27
Progress-Claim Integrity
Compares the physical percent complete claimed for an activity against the share of its duration actually consumed and the remaining duration it still carries.
Progress that is not reflected in the work the schedule still has to do is either an inflated claim or a remaining duration nobody drew down. Both change the forecast.
AACE RP 29R-03 §3.3.E, SCL Protocol §11
Rewritten Actuals
Between updates, flags a recorded actual start or finish that changed after it was first reported, and percent-complete values that contradict the recorded remaining duration.
A recorded actual is a historical fact. One that moves between submissions undermines every date downstream of it.
AACE RP 29R-03 §3.3, SCL Protocol §11
Critical-Path Churn
Measures how much critical-path membership turns over between updates, and flags an activity that leaves the critical path in the same update its forecast finish slips.
A stable programme has a stable critical path. Membership that turns over as dates slip points to logic or constraint edits, not progress.
AACE RP 29R-03 §3.3, SCL Protocol §11
Milestone Deliverability
Lists the key milestones whose own total float is negative.
A reader sees the specific contractual dates the schedule itself forecasts to miss, rather than a bare count of negative-float activities.
DCMA 14-Point §7, AACE RP 29R-03 §4.3, SCL Protocol §8.1
Each flag reports evidence, never intent. The same movement can come from an honest re-plan, a data-entry correction or a deliberate presentation choice, and the file cannot tell them apart. That judgement stays with the reader.
Reference Standards
The published standards that inform our analysis checks.
| Standard | Full Title | Scope |
|---|---|---|
| AACE RP 29R-03 | AACE International Recommended Practice 29R-03: Forensic Schedule Analysis | Delay analysis methodologies, schedule quality indicators |
| DCMA 14-Point | US Defense Contract Management Agency 14-Point Schedule Assessment | Schedule quality metrics and thresholds. Every failing item lists its offending activities (full lists in the Excel appendix). On tracked projects the contested numeric limits can be tuned to the project — a tuned scorecard is always labelled “(custom thresholds)” and never presented as the canonical DCMA assessment; zero-tolerance checks are not configurable. |
| PMBOK Guide | PMI — A Guide to the Project Management Body of Knowledge (PMBOK® Guide), 8th Edition, 2025 | Schedule Performance Domain (§2.3) |
| PMI Practice Standard | PMI — Practice Standard for Scheduling, 2nd Edition, 2011 | Schedule model construction, network logic, calendars |
| CIOB Guide | CIOB Guide to Good Practice in the Management of Time in Major Projects | Schedule management best practices |
| SCL Protocol | Society of Construction Law Delay and Disruption Protocol, 2nd Edition | Delay analysis, concurrent delay, EOT assessment |
Bounded What-If Scenarios
On tracked projects, ScheduleLens can answer one bounded question about a retained snapshot: what is the network consequence of a single change — an activity taking N more working days, or a duration restored to its prior value? The engine calculates that consequence and reports the difference. Results are always movements relative to the schedule's own base, never absolute dates, and never a prediction of what will happen.
The model accounts for
- The schedule's own calendars, logic and lags
- Activity constraints
- Data date, actuals, and in-progress remaining work
The model does not include
- Some constraint types (e.g. as-late-as-possible) and forced finish dates
- Imposed project finish dates
- Resource levelling
- Sub-day precision (durations are whole working days)
Every scenario result carries this disclosure: it is a bounded scenario for one change, not a measurement or a forecast — on complex schedules, treat it as indicative.
What We Don't Do
- We do not overrule your scheduling tool. Float and critical-path membership in a report are the values your own software calculated. We run an independent CPM alongside it, but only to flag a disagreement (a stale critical path) and to measure change between snapshots — and a figure from our engine is labelled as ours.
- We do not assess contractual entitlement or determine fault.
- We do not replace the judgement of a qualified delay analyst.
- Reports are analytical output, not expert opinion or legal advice.
- We recommend professional review before relying on findings for formal submissions.
References to published standards are bibliographic citations indicating the analytical basis for each check. ScheduleLens is not affiliated with, endorsed by, or certified by AACE International, the Project Management Institute, DCMA, the Chartered Institute of Building, or the Society of Construction Law.