Turning a Door Schedule Into a Priced Division 8 Submittal Before the GC Moves On

Pricing doors and frames for Division 8 requires reconciling each opening across the door schedule, floor plans, elevations, partition schedule, and 087100 hardware specification before assigning a price. The schedule links those documents, but they still need to be read. Producing a defensible priced submittal before the GC moves on therefore requires reconciling the documents.

Why the door schedule is a bridge, not a source

A door schedule row identifies an opening through its door number, size, material, frame type, fire rating, glazing requirement, and hardware set number. Those fields make the schedule a useful navigation index. They do not make it a complete takeoff. Some information sits elsewhere by design, and a row can look complete while leaving essential pricing questions unanswered. For commercial hollow metal doors and frames, as for other architectural openings, the scheduled description is only part of the required information.

Handing comes from the floor plan, while leaf count comes from the door elevation. Wall type requires the partition tag to be cross-checked against the partition schedule, and frame throat cannot be resolved from a schedule that omits the necessary wall information. Each document supplies a different part of the opening. Without those references, the estimator has quantities but still lacks some of the details that show what each opening needs.

The hardware set number creates the same dependency. The actual hardware content lives in the 087100 spec, so extracting the set number does not establish what must be furnished or priced. The schedule bridges the architectural drawings and the hardware specification, but neither side can be resolved from the bridge alone.

Where a count that reconciles against the schedule is still wrong

The opening count needs an independent check against the plans. The counting relationship is Verified Openings = Plan Count + Schedule-Only Units − Duplicates − Exclusions. This makes the floor plans a parallel source rather than a document consulted only when a schedule entry looks unusual. Matching the schedule total is not enough.

Three failures can coexist on the same job. Doors drawn on plans may never appear in the schedule, scheduled doors may have no corresponding location on a plan, and doors may appear multiple times in plans but fewer times in the schedule. These discrepancies do not resolve one another. An over-count in one area can hide an under-count elsewhere, so the total looks right even though it counts the wrong openings. A recount against the same schedule preserves that error.

Pairs make the distinction between openings and leaves concrete. The schedule may or may not identify an opening as a pair, while the elevation shows both leaves drawn. Pair status and leaf count must therefore come back from the elevation into the takeoff. The opening can be present in the count while its leaf quantity remains wrong. This is a cross-document error before any hardware set has been examined.

The two hardware set conflicts that never move the opening count

Hardware conflicts are harder to detect because they leave the opening total untouched. The first conflict occurs when the schedule and specification assign the same door to different hardware sets. Both references can look valid in isolation. Only the comparison reveals that the opening has two competing definitions, and the estimator must decide which document governs. Counting that door again provides no help.

The second conflict arises when a set number in the schedule is nowhere in the hardware specification. The opening remains visible, and the schedule still supplies what looks like a hardware reference. There is simply no corresponding bill of materials. A check of totals can pass that opening with the hardware still missing because it checks only whether the door is there.

Where finish designations are absent from a significant share of hardware line items, including gasketing, thresholds, door sweeps, and silencers, the estimator is pricing assumptions. A valid set reference does not resolve that omission. The set can be found and its components extracted while the information needed to price those components remains incomplete.

These failures occur because a sequential process can finish the schedule, move to the specification, and never compare the two assignments at the opening level. An omitted hardware line on an exterior door that goes uncorrected during takeoff then becomes a change order on a live project. By that point, award has passed and schedule pressure is highest.

How institutional owner standards change what the spec actually asks for

Institutional projects bring a third source into the hardware check. The project specification may reference an owner standard that independently prescribes brand, grade, or format. In that case, the project spec is not the final word. Pricing its generic requirements without the referenced standard produces a bid against an incomplete set of requirements, even if every scheduled opening has been matched to a hardware set.

Michigan DTMB’s door hardware specification, last revised April 30, 2025, permits any manufacturer belonging to a hardware manufacturers’ trade association for some categories, including hinges and thresholds. It mandates specific sole-source brands for others, including Allegion-Von Duprin for exit devices and Dormakaba-Best for locksets. The distinction changes what can be priced. The same standard also prescribes mechanical details down to cylinder diameter, pinion dimensions, and hydraulic fluid temperature range. Those requirements appear in the owner document.

Michigan State University publishes no numbered hardware sets in its specification. Instead, it organizes hardware selections by component type across tables, requiring the estimator to match each opening’s usage to the correct row and repeat that work for every opening. There is no set reference to anchor the bill of materials. The estimator is assembling hardware sets against the owner’s rules rather than extracting ready-made sets from the project documents.

That format changes the task at its foundation. A schedule-to-spec comparison cannot finish the takeoff when the hardware definition depends on a separate owner table. The owner standard must be a parallel input from the start. Otherwise, the later review revisits decisions already made against requirements that were never complete.

What sequential document reading costs in real time

Every return to a document consumes time that cannot go to the next quote. Sequential reading creates those returns whenever information extracted from one source needs to be reconciled with another. The cost compounds as bid volume rises. A fast first pass through the schedule offers little advantage if handing, pairs, wall types, and hardware assignments all require separate trips back through the set.

At a distributor, a salesperson often performs the takeoff between calls. Quote requests then get sorted by how long they take, regardless of what they are worth. Complex institutional packages are especially exposed to that constraint because their hardware requirements demand more reconciliation. The largest jobs can wait while the simpler requests get done. The distributor cedes that work by default, even without deciding that it is unattractive.

Custom frames, specialty glazing, and electrified hardware can have lead times of weeks. Finding a missed item after award can thus cause trouble both with buying it in time and with paying for it. The project is already running when the estimator identifies the missing requirement. The delay is not confined to the time needed to revise the takeoff.

Pricing itself has a short useful window. Distributor quotes are typically valid only briefly, while material costs for brass, heavy steel, aluminum mill shapes, and steel mill products have been moving week to week in 2026. Time spent reconstructing an opening’s requirements reduces the time available to price it while the quote remains usable.

Why pricing doors and frames in Division 8 requires simultaneous document reading

The structural fix is to reconcile the documents as each opening is processed. Handing is read against the floor plan, leaf count is established from the elevation, wall type is pulled through the partition schedule, and the hardware set is checked against the 087100 spec. On institutional work, the owner standard participates in that same comparison. This is one cross-document pass rather than a chain of isolated extractions followed by returns.

A schedule can be internally consistent, and a specification can be internally consistent, while the two disagree about the same door. Neither isolated reading exposes the contradiction. A sequential process that eventually catches it does so by returning to compare the documents.

The prescription does not eliminate the need for judgment. Contradictory specifications, undefined finishes, contested conventions, and cross-division scope questions remain unresolved by extraction alone. Electrified hardware can raise questions about who owns the power supply and how to coordinate with Division 26 and 28. Having all the documents at hand shows where someone needs to decide who is responsible, but leaves that decision open.

A hardware assignment that remains unresolved should be flagged as a conflict so that a price that looks complete does not hide it. The estimator’s effort can then go to deciding what governs, what belongs in scope, and what still needs pricing.

What AI-assisted takeoff actually automates and where the estimator still decides

AI-assisted Division 8 takeoff lets the documents be read together fast enough to meet a GC’s timeline. The strongest platforms extract door schedules, hardware sets, and floor plan data together, map each opening to its hardware set, and flag disagreements between scheduled assignments and specification definitions. The automation works across the relationships that manual sequential reading can leave unexamined.

It also fills missing schedule columns from the corresponding documents. Handing comes from the floor plan, leaf count from the elevation, and wall type from the partition schedule. A completed row is useful only when its contents reflect the relevant documents.

The estimator still decides which document governs conflicting hardware assignments and what finish to use when the code is blank. Sets marked out of scope or bearing an OT designation also require a scope decision. Pricing remains incomplete where the spec requires an item that no manufacturer quote has covered. Those are not counting tasks. The tool can expose the gap, but the estimator must determine how the bid handles it.

Electrified hardware makes that boundary especially clear. Card readers, electric strikes, magnetic locks, power transfers, request-to-exit devices, door-position switches, and power supplies can each bring Division 26 and 28 coordination questions. Extraction shows what the item is but leaves open who is responsible for it. The estimator resolves responsibility so that a detected component does not become either an omitted requirement or an unsupported inclusion.

General-purpose opening-count tools do not close this gap when they price hardware at the line-item level without reading the specification against the schedule. Set conflicts, missing set numbers, and out-of-scope designations remain invisible under that approach. A platform built for Division 8 knows the document types, terms, and links between them that it needs to reconcile each opening. Its value ends where the estimator’s judgment begins, with the unresolved decisions visible before a price is committed.