Spare Parts Inventory Management System for Thailand Factories

A spare parts inventory management system for a factory in Thailand must do more than digitize receipts and issues. It must connect production-stop risk, supplier lead time, approved substitutes, repairability and accounting treatment in one decision process. This guide turns that objective into concrete requirements for equipment BOMs, maintenance plans, reorder points, alternatives, lot and serial tracking, repair returns, cycle counts, permissions, acceptance testing and a 90-day proof of concept.

The problem is not inventory value alone

Demand for raw materials and finished goods is often planned with forecasts, replenishment lead times, service levels and turnover. Maintenance, repair and operations (MRO) parts behave differently. Demand can be intermittent, a failure can create an immediate need, and the operational consequence of the same quantity shortage varies widely. One inexpensive fuse may be the only missing item needed to restart a critical machine, while several expensive motors can remain untouched for years.

The purpose of maintenance-parts inventory management is therefore not simply to make the book quantity match the shelf. It must support four outcomes at the same time:

1. Supply the right part to the right asset, work order and location when it is needed.
2. Avoid inventory that is not justified by downtime and supply risk.
3. Distinguish genuinely usable stock from reserved, suspect, defective or repairable items.
4. Give maintenance, production, purchasing, stores and finance a shared evidence base.

There is a real trade-off. Maintenance can reasonably prefer more protection, while finance can reasonably question slow-moving stock. A system should not automate one department’s preference. It should expose the cost, risk and performance consequences so that accountable people can approve a balanced policy.

The public ISO/TC 251 explanation of ISO 55001:2024 highlights decision-making and value, risk and opportunity, data and information, lifecycle management and predictive action. The standard does not prescribe a spare-parts stock level. Its public framing nevertheless supports treating a spare part as part of an asset-management decision, not merely as a quantity on a shelf. The article uses that logic as a design reference and does not make a certification claim.

Classify asset criticality and supply risk separately

Applying one replenishment rule to every SKU can produce two failures at once: shortages of critical items and excess stock of low-impact items. Start by separating asset criticality from part supply risk, then combine them to set the intended service policy.

Asset-criticality questions

• Could failure affect safety, the environment or a legal obligation?
• What production and recovery loss follows from downtime?
• Could the failure create quality or traceability problems?
• How many upstream or downstream processes would be affected?
• Is there another machine, manual method, subcontractor or controlled workaround?

Supply-risk questions

• What are the normal lead time and its variation?
• Is the item imported, subject to customs, a minimum order or special transport?
• Is there one maker, an end-of-life notice or knowledge held by one person?
• Can a locally sourced or technically equivalent item be approved?
• Can the failed unit be repaired, and how reliable is the repair turnaround?
• Does storage require an expiry date, humidity control, calibration or preservation work?

For example, an organization may combine A/B/C equipment criticality with high/medium/low supply risk and reserve its highest protection for A × high items. The labels and thresholds are local policy, not a universal benchmark. Price alone does not determine criticality, and consumption frequency does not determine it either. A rarely used insurance spare can be essential; a frequently used commodity may be available on the same day.

Translate the matrix into explicit stocking policies

The system should store the reason, approver, approval date and next review date—not only a score. Two parts can have the same score for different reasons, and a future change in supplier, equipment or production plan may affect them differently.

Build an equipment BOM that maintenance can actually use

A common failure is the separation of the item master from the asset register. Staff can see that a part exists but cannot reliably determine which machine, variant or revision can use it. SAP’s public maintenance BOM help describes the relationship between a maintenance object and the parts relevant to maintaining it. Whatever software is chosen, managing this relationship is foundational.

An operational equipment BOM should cover:

• Plant, area, line, asset, assembly and component hierarchy.
• Manufacturer part number, internal item code, previous number and successor number.
• Standard quantity per asset and installation position.
• Compatible equipment model, serial range and engineering-change level.
• Primary, permanently approved and temporary substitute relationships.
• Required tools, calibration, work instruction and safety-procedure links.
• Replacement interval, condition-monitoring point and failure-mode relationship.

A record stating only that “bearing B fits machine A” is unsafe if a later engineering revision no longer accepts the old specification. Add effective dates, asset serial ranges and technical approval. When a work order reserves parts, it should read the effective BOM version for that asset.

Resolve duplicate item codes without destroying history

The same component is often registered more than once because of spaces, hyphens, uppercase/lowercase, Thai/English descriptions, distributor codes or previous manufacturer numbers. Do not merge records by deleting them immediately. Create a normalized comparison key, identify candidates, inspect the physical item and specification, and let maintenance and purchasing approve the primary code and searchable aliases.

Before consolidation, check open purchase orders, reservations, repair orders, shelf balances and financial records. Freeze new transactions on the old code while preserving its history and alias. Master cleanup and transaction-history preservation are separate workstreams.

Separate physical, available, reserved and repair quantities

Ten units on the books do not necessarily mean ten units can be issued now. Oracle’s public field-service-parts overview distinguishes on-hand, available and reserved quantities, and also surfaces excess or shortage. A factory does not need to copy any product’s terminology, but it does need an equally clear state model.

An illustrative inventory-position formula is “available + reliable open supply + qualified repair returns − reservations − unresolved demand.” If repair dates are frequently missed, counting every planned repair return as firm supply will hide shortage risk. Make inclusion rules depend on date confidence and vendor evidence.

For the wider systems landscape, see our inventory management system implementation guide for Thailand. For barcode and point-of-work transactions, see the handheld terminal implementation guide.

A reorder-point system needs more than average usage

Reorder points are commonly based on lead-time demand plus safety stock. With maintenance parts, however, demand may remain zero for months and then appear suddenly. An average alone can understate the consequence of a shortage. SAP’s public MRO planning material discusses safety stock and reorder points, but each plant must set the values using its own data and risk decisions.

Select a replenishment method by item behavior

1. Reorder point: consumables with reasonably observable demand and lead time.
2. Min-max: items replenished at a periodic review or scheduled delivery.
3. Two-bin: low-value, frequent-use items kept near the work area.
4. Insurance spare: protection justified mainly by downtime consequence rather than demand frequency.
5. Non-stock: an item that can be ordered when required because delay or an alternative is acceptable.
6. Repair loop: a rotating population of serviceable, installed, failed and in-repair serialized units.

Service-parts planning must also see the network. A central warehouse, factory store, line-side cabinet and technician stock can show excess and shortage simultaneously if transfers and transfer time are ignored.

Worked reorder-point example

The following is fictional and only demonstrates a calculation. It is not a market benchmark or recommendation.

• Item: a sensor used for scheduled replacement and unexpected failure.
• Reviewed planning demand: 12 units over 90 days.
• Procurement lead time: 45 days.
• Planned lead-time demand: 12 × 45 ÷ 90 = 6 units.
• Safety stock approved by the plant: 3 units.
• Reorder point: 6 + 3 = 9 units.
• Available stock: 7 units.
• Reserved stock: 2 units.
• Reliable open purchase supply: 1 unit.
• Inventory position: 7 + 1 − 2 = 6 units.
• Gap to reorder point: 9 − 6 = 3 units.
• Order multiple: 4 units.
• System proposal: order 4 units.

Looking only at the seven physical units can create false confidence. After reservations and reliable incoming supply, the position is six. The proposal rounds the three-unit gap to the approved order multiple of four. A real calculation may also consider planned maintenance, cancellable reservations, delivery reliability, common-cause failures, minimum orders and shelf life.

Store the reason with the number: demand window, lead-time observation date, safety-stock approval basis and review date. Require a reason for overrides. Our factory stockout prevention system guide expands on exception signals and operational response.

Manage substitutes as conditional technical approvals

A single “substitutable” flag is inadequate. Mechanical fit, electrical compatibility, control-program support and quality approval are different tests. Relationships can also be directional: B may replace A, while A cannot replace B.

Each substitute relationship should state:

• Original and substitute item.
• One-way or two-way applicability.
• Equipment, model and serial range.
• Required machining, adapter, wiring or parameter change.
• Safety, quality or customer-approval condition.
• Temporary or permanent status and expiry date.
• Technical approver, evidence document and validation result.
• Whether the asset must later be restored to the original design.

When a shortage occurs, display only substitutes valid for the selected asset and work conditions. If a temporary substitute is installed, automatically create a follow-up action and track it until permanent disposition.

Choose lot or serial traceability by item

Serializing every bolt imposes unnecessary effort; tracking nothing can make a defect or warranty problem impossible to contain. Choose the traceability level based on value, repairability, safety, warranty, calibration and quality impact.

Lot tracking commonly fits lubricants, adhesives, chemicals and seals with shelf lives, or items requiring batch-level defect containment. Serial tracking commonly fits servo drives, PLCs, instruments and motors whose individual repair, warranty, calibration, firmware or movement histories matter.

At issue, capture who issued which lot or serial, when, for which work order and asset. At return, distinguish unused return, defective return and mistaken issue. A practical scanning flow identifies the asset, location, work order and part through barcodes or QR codes. The aim is not maximum data entry; it is sufficient evidence at the point where the event happens.

Make the repair-return loop visible

When repairable failures do not return to the store, the plant can buy unnecessary new units. Track a repairable item through an explicit flow:

1. Remove it from the asset and record the symptom, asset and work order.
2. Return it to a quarantine location, never directly to serviceable stock.
3. Decide warranty, repairability, quotation limit and disposal authority.
4. For external repair, track shipment, receipt, quotation and expected return.
5. Inspect, test or calibrate the returned unit; only a pass can restore availability.
6. Store repair cost, failure mode and time to recurrence in the serial history.

Do not mix “in repair” with “serviceable.” Measure turnaround and repair outcome, while retaining the date and confidence of a promised return. Vendor changes, parts shortages and end-of-life conditions can invalidate past performance.

Repeated failures or cumulative repair cost above an approved rule should trigger a repair/replace/dispose review. The threshold is organization-specific and should be an authorized configuration, not a hidden hard-coded number.

Replace annual correction with continuous inventory accuracy

Slow-moving parts can conceal errors for long periods. Combine annual counts with risk-based cycle counting. Prioritize critical, high-value, expiry-controlled and historically inaccurate items. Use blind counts when appropriate, separate first count from recount, classify root causes, and require approval for adjustments.

Do not use an inventory adjustment to hide a transaction that should be corrected. Control movements during a count or reconcile them by timestamp. Measure differences by criticality, quantity, value and recurrence. One missing insurance spare can matter more than many discrepancies in inexpensive fasteners.

Put permissions and audit trails at the center of the RFP

The key question is not only what a user can enter, but who can change a rule and how the decision can later be explained. Separate at least these duties:

• Item creation and master approval.
• Technical approval of substitutes.
• Reorder-point or safety-stock change and approval.
• Purchase request, purchase approval and receipt.
• Receiving, inspection and release to available stock.
• Issue, return and inventory adjustment.
• Repair order, disposal decision and disposal approval.

An emergency issue may be necessary. Rather than banning it, require a reason, asset, work order, approver and post-event review deadline. Personal IDs are the default; even a shared terminal must identify the user.

An audit trail should retain before value, after value, time, operator, approver and reason. Define who can read logs and how long they are kept. Test backdating, negative stock, incomplete transfers and reversals during acceptance.

Do not let a system label determine accounting treatment

IAS 16 paragraph 8 states the principle that spare parts, stand-by equipment and servicing equipment are recognized under IAS 16 when they meet the definition of property, plant and equipment; otherwise they are classified as inventory. The name “spare part” therefore does not put every item into one accounting category.

Finance and, where relevant, auditors must determine the accounting policy. The implementation team should not invent it. Keep operational location and usability separate from financial classification, valuation, impairment and disposal, while maintaining the keys needed for integration.

Holding a long-lead insurance spare is an operational risk decision. Deciding how to recognize or depreciate it is related but not identical. Avoid an RFP promise such as “automatically compliant with accounting standards.” Ask how the system supports the organization’s approved policy through data, workflow and posting integration.

Requirements to include in a spare parts system RFP

Business scenarios and decision rules are more useful than a list of screens.

Functional requirements

Non-functional and implementation requirements

• Thai, English and Japanese display, Unicode search and safe CSV import/export.
• Offline or retry control in factory areas with unstable Wi-Fi.
• Barcode/QR scanners, printers and handheld-terminal integration.
• APIs or batches to ERP purchasing/accounting, CMMS/EAM, MES and identity systems.
• Role control, audit log, backup, recovery and retention.
• Transaction volume, concurrency and performance during month-end or counts.
• Test environment, migration rehearsal, training, Thailand support and incident process.

If a supplier says “standard function,” ask for the configuration, limits, cost and upgrade impact. If it says “customizable,” ask who supports the customization and how it is retested after upgrades. Demonstrations should include wrong units, duplicate scans, expired parts, reservation conflicts, communications loss, returns and reversals—not only a clean normal flow.

Acceptance tests should reproduce a real stoppage response

Use representative assets, parts and realistic work orders for end-to-end acceptance.

1. Register an asset failure and find valid parts and substitutes from the effective equipment BOM.
2. With two available, one reserved and one in inspection, confirm what another job may issue.
3. Let a critical part cross its reorder point and verify the order multiple and approval path.
4. Exclude a substitute valid for an old asset but invalid for the newer revision.
5. Issue a serialized unit to the failed asset and move the removed unit into repair.
6. Fail inspection on a repaired return and verify that it does not become available.
7. Confirm that a counter cannot approve their own material inventory adjustment.
8. Retry transactions after a communications interruption without duplicates.
9. Change safety stock and verify before/after value, reason and approver.
10. Cancel or correct an ERP order and verify the inventory position.

Pass criteria must define the expected quantity, state, permission, integration result and audit event—not just that a page appeared. Classify defects by severity and agree who approves a temporary workaround and production go-live.

Use a 90-day PoC to validate both system and operating model

A PoC is not a product demonstration. It should test implementation assumptions in a controlled scope. Choose one line or asset group with meaningful downtime impact, accessible history and committed local owners. Do not spread the effort across every item and every plant.

Days 0–30: establish the baseline and usable data

• Confirm assets, items, stocking locations, roles and success criteria.
• Review duplicate codes, units, equipment BOMs, states and open orders.
• Approve criticality rationales.
• Connect scanners, terminals, network and selected ERP/CMMS interfaces.
• Record initial physical/book differences separately from later improvement.

At the Day-30 gate, review scope coverage, initial inventory accuracy, missing masters, user permissions and interface errors. Where a baseline was previously unknown, do not claim improvement; measurement capability itself is the first deliverable.

Days 31–60: run normal and exception transactions

• Execute planned replacement, unexpected failure, receipt, emergency issue, return and repair.
• Compare the system’s reorder proposal with the planner’s approved decision.
• Test reservation conflict, substitute, failed inspection and communications interruption.
• Monitor transaction delay, unposted work, difference reasons and alert response.
• Improve Thai/English training and shift handover.

At Day 60, decide whether the workflow is sustainable in daily operations. If workers delay entry and return to spreadsheets, revisit the interface, terminal placement, responsibility and approval steps.

Days 61–90: validate replenishment and decide expansion

• Review lead time, reservations, repair returns and reorder assumptions using observed evidence.
• Review shortage, excess and slow-moving candidates together with criticality.
• Test permission audit, backup/recovery, month-end and cycle count.
• Define production migration, data owners, training, support and improvement backlog.
• Prepare a continue, modify, narrow or stop decision.

Do not promise that inventory value will fall or stockouts will reach zero within 90 days. A low-frequency failure may not occur in the period. The PoC should show whether decision data is obtainable, controls protect critical parts, operational effort is acceptable, and expansion cost and risk are explainable.

Illustrative PoC scorecard

These are planning examples, not universal thresholds. The PoC charter should state the formula, data source, owner, decision date and exception authority for every measure.

Common implementation failures

Trying to migrate every item before any use begins

A program can stall while trying to perfect every dirty record. Limit the first scope to critical assets and parts, and tag unverified data with an owner. This does not permit ambiguous safety or financial data; it makes the scope explicit.

Calculating and freezing every reorder point in one batch

Demand, downtime impact and repairability differ. Select methods by item, review exceptions and refresh observed lead time. Define how outliers are treated.

Posting shop-floor issues the next day

The system then overstates supply and can create conflicting reservations or emergency buys. Minimize input, connect scans to work orders, place devices at the work point and monitor transaction delay as a process problem.

Returning repairables directly to serviceable stock

A failed or untested return can cause another stoppage. Align the physical quarantine, status label, system state and quality decision.

Using inventory reduction as the only KPI

That can make a rarely used critical part the easiest target. Review equipment status, common use, substitutes, lead time, repairability and possible transfer or sale before disposition.

FAQ: selecting and implementing the system

What is a spare parts inventory management system?

It connects maintenance-part masters, equipment BOMs, availability states, reservations, replenishment, repair, substitutes, traceability and counts so the plant can manage downtime risk and carrying cost. It commonly integrates CMMS/EAM work with ERP purchasing and accounting.

Can maintenance parts inventory be managed in Excel?

Excel can help establish a policy when scope, sites and editors are limited. It becomes fragile when concurrent updates, reservations, serial history, permissions, audit trails, equipment BOMs and multiple locations must operate in real time. It can remain a cleansing tool, while the production system of record is explicitly defined.

Can a stockout prevention system eliminate stockouts?

It cannot guarantee that. Unexpected demand, supply disruption, customs, disasters and quality problems remain. Criticality, substitutes, reorder rules, safety stock, repair loops, site transfers and exception alerts reduce the most unacceptable risks.

Who should approve values in a reorder-point system?

The system can calculate candidates, but maintenance, purchasing, production and finance should review the evidence and approve through defined authority. Demand history is only one input alongside asset impact, lead-time variation, alternatives, repairability and shelf life.

How does the system differ from ERP or CMMS/EAM?

ERP commonly centers on purchasing, inventory and accounting; CMMS/EAM commonly centers on assets, maintenance plans and work orders. Product scope varies. Before adding another platform, identify what current tools provide, which master is authoritative, how interfaces are owned and where the shop-floor experience is inadequate.

What should a Thailand factory include in a 90-day PoC?

Select one line or asset group with meaningful impact, accessible parts and transaction history, and committed local owners. The scope should complete the loop from receipt and issue through failed return, repair, replenishment and count rather than sampling every item superficially.

Summary: manage recovery capability, not just quantities

The value of a spare parts inventory management system is not an electronic stock list. It connects parts to assets through an effective BOM, separates criticality from supply risk, distinguishes available, reserved, inspection and repair states, and makes replenishment and substitute decisions explainable. Lot and serial traceability, repair return, cycle counting, authorization and audit evidence let the same management discussion address downtime protection and excess stock. Describe real scenarios in the RFP, verify quantities, states, permissions and interfaces in acceptance, and use the 90-day PoC to establish measurable, sustainable operations without exaggerating short-term effects.

TOMAS TECH can help a Thailand factory map its maintenance, warehouse and purchasing process; define the equipment BOM and replenishment policy; design barcode operations and ERP/CMMS integration; and prepare acceptance tests. If you are still defining scope or PoC decision criteria, you can contact us at that early stage.

References

• ISO/TC 251, ISO 55001:2024 key changes: https://committee.iso.org/sites/tc251/home/projects/published/iso-55001.html
• ISO, ISO 55001:2024 overview: https://www.iso.org/standard/83054.html
• IFRS Foundation, IAS 16: https://www.ifrs.org/content/dam/ifrs/publications/pdf-standards/english/2022/issued/part-a/ias-16-property-plant-and-equipment.pdf
• SAP Help, Maintenance BOM: https://help.sap.com/docs/SAP_S4HANA_ON-PREMISE/e72f747389b340229f7fa343975bfa57/9d71b65334e6b54ce10000000a174cb4.html
• SAP Help, Service Parts Planning: https://help.sap.com/docs/SAP_S4HANA_ON-PREMISE/677f0a4e71d7487ebb70683014761789/47e12f019e014ac5e10000000a42189d.html
• SAP Help, MRO planning: https://help.sap.com/docs/SAP_INTEGRATED_BUSINESS_PLANNING/feae3cea3cc549aaa9d9de7d363a83e6/89ba7acf063047cea5b37702adfa300b.html
• IBM, Spare Parts Management: https://www.ibm.com/think/topics/spare-parts-management
• IBM Maximo, Inventory module: https://www.ibm.com/docs/en/max-it/cd.0.0_cd?topic=applications-inventory-module
• Oracle, Managing Field Service Parts: https://docs.oracle.com/en/cloud/saas/service-logistics/26a/fasul/overview-of-managing-field-service-parts.html
• Thailand BOI, Thailand FastPass press release, 23 June 2026: https://www.boi.go.th/upload/content/PR95_2569EN.pdf

This article provides general systems-design information, not accounting, legal or safety advice. Applicable standards, accounting treatment, legal obligations and stock levels require review by accountable specialists. Every numerical value is fictional and illustrative, not a market price, customer result or performance promise.