Project controls forensics, as it has historically been practiced, is a single-stream discipline. The schedule analyst reviews the schedule. The cost analyst reviews the cost data. The contracts attorney reviews the contract. In large program disputes, these analyses eventually converge — but they converge late, in the context of a claim or litigation, after the events that generated the findings have long passed.
The multi-modal approach to project controls analysis treats the contract, the schedule, and the cost data as a single integrated dataset that must be read together in real time. The findings that emerge from this integration cannot be produced by any single stream alone. They represent a category of program intelligence that exists only at the intersection.
What Each Data Stream Contains Alone
A project schedule in isolation contains the program's temporal logic: what was planned, when work was performed relative to plan, how relationships between activities evolved, where float accumulated or eroded, and how the critical path composition shifted across reporting periods. It does not contain the financial implications of those events, the contractual rights they trigger, or the cost data needed to validate whether reported progress reflects real resource consumption.
A cost dataset in isolation contains the financial record of the program: actuals by cost element, earned value by control account, variance explanations, and EAC trends. It does not contain the schedule context needed to evaluate whether costs are being incurred on the scope that the schedule says is in progress, or the contract provisions that define what cost growth is reimbursable versus absorbed.
A contract document in isolation contains the rights and obligations of the performing organization: the scope baseline, the change order provisions, the notice requirements, the float ownership language, the liquidated damages thresholds, and the dispute resolution procedures. It does not contain the performance data needed to evaluate whether those rights have been triggered or those obligations violated.
Each stream is rich. Each stream is incomplete. The program intelligence that drives real decisions lives in the connections between them.
The Schedule-Contract Connection
The most immediately actionable cross-modal finding is the identification of contractual notice obligations triggered by schedule events. As discussed in prior coverage of the notice window problem, capital project contracts require written notice as a precondition to claims for changed conditions, differing site conditions, and scope growth. The triggering events for those notice requirements are visible in the schedule — actual start dates, predecessor delays, milestone misses — but the notice requirements are in the contract.
A system that reads both simultaneously can identify, at the moment a schedule event is recorded, whether that event constitutes a compensable occurrence under the contract language, what the applicable notice period is, and how many days remain before the window closes. This is the real-time version of what contract attorneys reconstruct retrospectively during claims analysis — and performing it in real time is the difference between capturing the claim and forfeiting it.
The same connection surfaces float ownership disputes. Many construction contracts contain language specifying whether float belongs to the owner, the contractor, or the program — and whether a contractor can be directed to consume contractual float before damages accrue. Reading float position in the schedule against float ownership language in the contract produces a finding that neither stream could generate alone: whether the contractor's current float position is actually at risk or whether contractual float ownership provides a buffer.
The Schedule-Cost Connection
Cross-referencing schedule and cost data produces findings about whether reported progress is consistent with actual resource consumption. A control account that reports 70% schedule progress while showing cost accumulation consistent with 45% completion is exhibiting a pattern that raises questions about either the progress reporting or the cost tracking. Either the schedule is overstating progress or costs are being deferred — both of which are worth investigating.
The specific cross-modal check that is most resistant to evasion is the validation of physical progress claims against procurement and material data. A construction schedule that reports significant installation progress while procurement records show that materials have not yet been delivered to site is showing an impossibility. The schedule is reporting work that cannot yet have been performed, because the resources required to perform it are not present.
This cross-validation is the mechanism that the most sophisticated manipulation schemes cannot easily defeat. A contractor can simultaneously misrepresent schedule progress and cost accumulation in consistent ways. It cannot simultaneously misrepresent schedule progress and the third-party records of material deliveries, subcontractor invoicing, and inspection documentation.
The Cost-Contract Connection
The cost-to-contract cross-reference produces findings about whether cost growth is occurring in ways that are compensable under the contract versus in ways that represent contractor-absorbed overruns. A program with significant cost growth on a fixed-price scope area is accumulating overrun that no contract provision will recover. A program with equivalent cost growth on a cost-reimbursable scope area, resulting from owner-directed changes, may have substantial claim value that is not being captured.
This distinction requires reading cost variance explanations against contract scope boundaries. A variance analyst who attributes cost growth to "design complexity" without specifying whether that complexity resulted from owner-issued changes or contractor scope interpretation is producing an explanation that cannot be evaluated for claim value. The multi-modal system connects the cost variance to the specific contract provisions that define whether the underlying cause is compensable.
Integration as Architecture
The multi-modal approach is not a data aggregation exercise. It is an architectural requirement: the system must have ingested all three data streams at sufficient granularity to make the cross-references at the level of individual schedule activities, individual cost elements, and individual contract clauses.
A system that ingests a summary-level schedule, a roll-up cost report, and a contract abstract cannot make the real-time connections that produce actionable findings. The notice window finding requires knowing which specific activity's actual start date triggered which specific contract clause with which specific notice period. The procurement cross-validation requires comparing specific schedule activity progress claims against specific material delivery records for specific material types.
This is the granularity level at which the multi-modal approach produces findings that change program outcomes. Above that level, it produces reporting. At that level, it produces intelligence.
The Forensic Intelligence Engine ingests contract language, schedule data, and cost records as an integrated dataset — producing findings at the intersection that no single-stream analysis can generate.