A green balance can still hide unusable stock

The inventory view says 47 units are on hand. One operator has a container in front of them. Its label is legible, the material name is correct, and the printed expiry has not passed. But part of the lot is reserved, the storage unit has an open excursion, and this container is released for development work—not the GMP batch about to consume it.

The number is not wrong. It is answering the wrong question.

The operational question is not “How much do we have?” It is “Which exact containers may be used for this exact operation now, and what evidence supports that answer?”

Seal controls the physical population behind the balance: material, lot, container, quantity, place, quality state, environmental history, demand, and actual use. Each answer can be traced to the records and rules that produced it.

Purchasing may know what was ordered. A warehouse system may know what was received and moved. Quality knows what was released or held. Laboratory and manufacturing records know what was sampled, prepared, dispensed, or consumed. Local trackers often fill the gaps between them.

Each source can be accurate for its own job while the combined inventory picture is already stale. Seal keeps the distinctions explicit:

Production planning, laboratory scheduling, and point-of-use selection consume different views of the same controlled population—not separately maintained totals.

A material definition says what an item is. A lot preserves its manufacturing and supply identity. A container is the physical object that can be received, labeled, moved, opened, sampled, partially consumed, damaged, quarantined, or destroyed.

Instead of “polymerase, 47 units,” Seal can identify container C03 from supplier lot RM-0417: a 250 µL aliquot in ULT-03, shelf 2, box C, with 180 µL remaining, an intact seal, a current quality decision, and a complete movement and use history.

Pallets, drums, bottles, vials, bags, kits, plates, and aliquots can form parent-child handling units. A split creates new physical identities without losing source genealogy. A relabel changes the label, not the history. Current quantity and state are derived from what actually happened to each object.

The effective material definition controls approved manufacturers and sites, supplier scope, grade, specification, unit, packaging, storage, shelf life or retest, sampling, testing, release, hazards, potency, and use restrictions. Version changes apply from a defined date without rewriting historical receipts.

Receiving preserves purchase and shipment references, manufacturer and supplier lots, internal lot, quantities, condition, certificates, dates, ownership, and initial location. Labels are generated from the governed record. The incoming population normally begins in a state that prevents unapproved use while sampling, testing, document review, and disposition remain open.

Supplier datasheets, specifications, safety data sheets, and certificates can be extracted into a source-linked change set. AI proposes values; a qualified person reviews, corrects, and approves them. The source remains attached, and no extracted value becomes approved master data merely because a model found it.

Received, quarantine, sampled, under test, released, restricted, reserved, held, rejected, expired, consumed, returned, and destroyed are operating states—not colored labels on a report.

Every state change retains scope, reason, effective time, evidence, conditions, authority, and approver. A lot-level hold can propagate to every child container. A damaged-container restriction can remain limited to one physical item. Release can permit R&D use while still blocking GMP manufacturing.

Expiry, retest, open life, preparation life, and use-by rules can update eligibility automatically. If governance permits an exception, it follows a separate authorized path with defined scope, rationale, controls, duration, and approval. There is no generic “continue anyway” action that silently turns a failed check into an accepted use.

Before a batch step, laboratory method, preparation, or other controlled activity can consume material, Seal evaluates the intended material and grade, manufacturer, supplier, lot attributes, specification, quality status, dates, storage and excursion state, location, reservation, available quantity, product, market, process, method, restrictions, and user authorization.

The decision returns eligible, conditional, or blocked with the rule and evidence that produced the result. Selection lists can exclude ineligible stock; a barcode scan confirms the physical container at the point of use. An accepted dispense records the exact lot, container, quantity, operator, time, and execution context while updating the remaining balance.

The execution record therefore preserves not only what was used, but why it was considered suitable at that moment.

Receipt, transfer, split, consolidation, sampling, preparation, dispense, consumption, return, spill, loss, adjustment, destruction, and shipment debit and credit identified stock. The current balance is the sum of transactions—not an overwritten cell.

Expected and actual quantities, units and conversions, measurement basis, tolerances, discrepancies, reasons, witnesses, and approvals remain attached to the event. Container history shows who moved or changed it, when, from where, to where, and because of which activity.

Physical location can resolve from site to building, room, storage unit, shelf, rack, box, and position. The address remains useful to operators while the movement history supplies the evidence needed for custody, counting, and environmental impact.

A freezer alarm should not create a guess based on where inventory sits now. Seal combines the location hierarchy, container movement intervals, sensor coverage, event start and end, mapping, conditions, and alarm evidence to determine which containers were present during the affected period.

A container moved into the unit halfway through the excursion has a different exposure from one stored there throughout. Assessment, supporting stability or shipping evidence, status actions, and final disposition remain connected to the exact population. The same trace can identify work that already consumed an affected container.

On-hand inventory does not show what tomorrow's work has already claimed. Reservations connect production orders, laboratory plans, studies, campaigns, kits, shipments, safety stock, priority, quantity, expiry, and due time to eligible stock.

Planners can distinguish released inventory from available-to-promise inventory and see shortages before execution begins. FEFO rules can propose the appropriate containers without taking material reserved for higher-priority or earlier work. Substitution and reallocation remain explicit decisions with impact and approval.

Laboratory and manufacturing workflows record material at the point of sampling, preparation, dispensing, or consumption. The inventory transaction and the execution record refer to the same physical container and quantity, so reconciliation does not depend on someone posting a second summary later.

From a supplier lot or container, Seal can traverse samples, tests, preparations, experiments, batches, intermediates, finished products, shipments, retained samples, remaining stock, returns, and destruction. From a batch or result, it can trace backward to the exact containers, approvals, locations, excursions, and movements that supported the work.

When a supplier notification arrives, the question is no longer “Where might this lot have gone?” The trace separates stock still on hand, material already consumed, affected work in progress, released outputs, and ruled-out populations—with the evidence for each conclusion.

Cycle counts and full inventories freeze a defined population while accounting for transactions in progress. Counts can be blind, risk-based, location-based, material-based, or sampled. Operators record observed quantities and conditions; the system routes recounts, discrepancies, impact assessments, adjustments, and approvals.

An adjustment does not erase the difference. Expected quantity, observed quantity, cause, evidence, decision, and resulting transaction remain visible. Repeated variance by location, shift, material, transaction type, or user can become a process-improvement or investigation signal.

Seal does not need to replace the ERP that buys the material or the WMS that moves it through a warehouse. Those systems can continue doing the jobs they already do well. Seal supplies the operational answer they usually cannot provide alone: whether this particular container is suitable for this particular use right now.

When a laboratory or manufacturing step requests material, Seal evaluates the live inventory state and records the actual consumption. When a storage monitor reports an excursion, Seal resolves the affected containers and downstream uses. Each system remains authoritative for its own record; the handoffs no longer leave eligibility or genealogy to inference.

Choose one material class with costly shortages, frequent expiry, difficult storage, or weak traceability. Govern its suppliers and storage rules, label new receipts at container level, and connect one real laboratory or manufacturing use. That is enough to test whether the digital record matches physical practice.

Before expanding, put the flow under pressure. Hold a supplier lot. Trigger a storage excursion. Create a count discrepancy. Trace one source lot forward and one batch backward. If the system can explain what exists, where it is, whether it may be used, what consumed it, and why the balance changed, the pattern is ready to scale.