MRP Says Available. Why Is the Kit Still Short?
Inventory quantity does not equal usable material. Allocation, location, quality, revision and timing determine whether a production kit is truly ready.
The short answer
MRP can show enough inventory while a production kit remains short because “on hand” is not the same as “usable for this job now.” Components may be reserved for another order, held by quality, stored at another site, assigned to the wrong revision, lacking customer approval or scheduled to arrive after the line needs them. Kit readiness must evaluate quantity together with allocation, location, condition, identity and timing.
For an EMS manufacturer, the solution is to connect MRP demand with warehouse, purchasing, quality, engineering and production status. The planner should see the exact blocking component and the reason it cannot be issued—not simply a green aggregate balance.
The comforting number that misleads production
Consider a job requiring 4,800 capacitors. The inventory screen shows 6,200 available, so the order is released. At kitting, only 3,900 can be picked. Another 1,000 are allocated to a customer programme with an earlier commitment. Six hundred are in incoming inspection. Seven hundred are physically stored at another plant.
The arithmetic was not wrong. The business meaning was incomplete.
Electronics material planning is especially sensitive to this difference because one missing low-cost component can stop a high-value assembly. Hundreds of line items may be ready, yet the kit is not executable. A percentage-complete measure can look excellent while the only relevant fact is that one critical reel is unavailable.
Availability has several dimensions
A credible material answer should address at least six questions.
Does the quantity physically exist?
Book inventory and physical inventory can diverge through delayed transactions, incorrect locations, scrap, unrecorded consumption or unit-of-measure errors. Cycle counting helps, but high-level accuracy does not guarantee that the specific component and lot can be found.
Is it free for this job?
Stock may already be allocated to another order, customer or programme. Reassigning it is a priority decision, not an automatic correction. The system should show the consequence of breaking the original allocation.
Is it usable?
Material can be on quality hold, expired, moisture-sensitive beyond its floor-life window, damaged or awaiting disposition. Treating quarantined stock as normal supply converts a material issue into a production or quality failure.
Is it the correct identity?
Similar descriptions are not proof of interchangeability. Manufacturer part number, approved vendor, customer restriction, package, value, tolerance and revision context matter. An alternate may be valid only under a controlled approval, as discussed in The Alternate Part Is Approved. But for Which Customer?.
Is it in the right place?
Inventory at a remote plant or external warehouse may be technically owned but unavailable for today’s setup. Transfer lead time, customs, inspection and receiving work must be reflected before that stock supports a start promise.
Will it be ready at the moment of need?
A purchase order due today is not the same as a reel at the line. Transport, receiving, inspection, storage and kitting create additional time. Planning should compare material readiness with the operation start, not merely the job completion date.
Why MRP and kitting often disagree
MRP is designed to calculate supply against time-phased demand. It depends on accurate master data and transaction timing. Kitting is a physical execution process. It encounters the actual bin, lot, label, package and condition.
The two disagree when the assumptions between them are not synchronized. Common causes include:
- BOM revisions taking effect without re-planning open jobs;
- allocations managed in a separate spreadsheet;
- quality status excluded from available-to-promise logic;
- substitute approvals stored in email;
- purchase dates reflecting supplier promise rather than usable receipt;
- consumption posted at shift end instead of at the operation;
- stock transfers represented without realistic lead time;
- engineering samples and production stock sharing the same balance.
The answer is not to distrust MRP. It is to improve the operational definitions and event discipline that feed it.
Build a material-readiness view by job
A planner needs a job-specific shortage view, not only an item-level inventory view. For every requirement, the system should connect:
- required quantity and need date;
- approved part and alternates for the exact product context;
- allocated quantity by location and lot;
- quality and expiry status;
- open purchase and transfer supply;
- remaining risk after expected receipts;
- owner and recovery action.
The result should classify shortages by cause. “Missing 900 pieces” is less actionable than “600 awaiting inspection until 14:00 and 300 allocated to Job B.” The second statement supports a decision: expedite inspection, split the job, reallocate stock or change sequence.
This is one component of the broader electronics production planning and control model. A job should enter the frozen dispatch window only when its material evidence meets the agreed readiness rule.
Avoid the kit-complete percentage trap
Kit completion is often expressed as a percentage of line items or total quantity. Both measures can hide the constraint.
If 199 of 200 items are ready, the dashboard reports 99.5 percent. But if the missing item is required at the first placement operation, the job is zero percent startable. Conversely, a component used only during final assembly may not need to delay the SMT start if replenishment is reliable and the risk is controlled.
Readiness should therefore consider operation sequence and criticality. The useful question is: can the planned quantity progress through the required horizon without an uncontrolled material stop?
Make shortage ownership explicit
A shortage list without ownership becomes a report that everyone reads and nobody resolves. Different causes require different owners.
Purchasing owns supplier confirmation and escalation. Warehouse owns location and picking discrepancies. Quality owns inspection and disposition. Engineering owns alternate approval and BOM correction. Planning owns sequencing and the decision to split or delay. Customer-program management may own a priority trade-off.
The system should record the action, responsible person, promised resolution and next review. When status changes, the schedule should recalculate or alert the planner. This replaces repeated meetings with a visible exception workflow.
Connect consumption back to planning
Material control does not end when the kit reaches the line. Actual consumption, attrition, scrap and returns affect the next job’s supply.
If feeders consume more than standard but the variance is posted later, MRP continues to promise stock that no longer exists. If unused reels return without timely location and condition updates, the system understates availability. MES and warehouse transactions should therefore close the loop while the schedule can still respond.
The Mars solution describes component planning, production scheduling, MES integration and real-time shop-floor visibility as connected capabilities. Their combined value is that material truth can follow the component from demand through issue, consumption and return.
A practical release rule
Before releasing a near-term job, verify:
1. The job uses the intended BOM revision and effectivity.
2. Every start-critical component is allocated in a usable status.
3. Approved alternates are valid for the actual customer and product.
4. Expected receipts include receiving and inspection time.
5. Remote stock includes transfer lead time.
6. Remaining shortages have an owner and accepted recovery plan.
7. The schedule reflects any quantity split or sequence change.
This rule need not eliminate every risk. It ensures that risk is visible before operators and machines wait for an answer.
Management should also review recurring shortage signatures. If the same component family is repeatedly found in the wrong location, the answer may be warehouse discipline rather than more safety stock. If incoming inspection is the recurring blocker, supplier quality and inspection capacity may need attention. Classification turns shortage history into a guide for structural improvement instead of a permanent expediting workload.
Frequently asked questions
Why does MRP show material available when the warehouse cannot pick it?
MRP may see aggregate on-hand or expected supply without accounting for allocation, physical location, quality hold, expiry, revision restriction or the time needed to receive and inspect material.
Should all components be kitted before a job starts?
Not always. The decision depends on operation sequence, replenishment reliability and risk. Start-critical items must be ready, while later-use components may follow under a controlled plan.
How should approved alternates appear in shortage planning?
They should be evaluated against the job’s customer, product, revision, site and effective dates. A generic approved flag is insufficient.
What is the most useful shortage metric?
Measure blocked jobs and start-critical shortages by cause, owner and age. A simple percentage of lines available can hide the one component that prevents production.
How does MES improve material planning?
MES can return timely consumption, scrap, completion and WIP events. Those signals improve inventory truth and help planners respond before the next job is affected.