Introduction

As global competition and the diversification of consumer needs accelerate, a shift to “High-Mix Low-Volume” (HMLV) production has become unavoidable for the manufacturing industry. To supply a wide variety of product variations in short cycles, even the slightest delay or inventory shortage can bring the entire production line to a halt.

Screws, in particular, which determine the final quality of a product, become extremely difficult to manage as the combinations of size, material, surface treatment, and strength classification increase exponentially. The dilemma of holding excess inventory for fear of stockouts, which pressures storage space and cash flow, versus cutting back too much and inviting line stoppages or increased costs from emergency procurement, is a major challenge that plagues on-site managers.

This article is written for managers in production control, purchasing, and quality control who face these issues. It explains practical know-how to achieve both “optimal inventory” and “cost optimization” simultaneously. We will cover the latest approaches in screw inventory management, from “visualization” using IoT and data analysis to demand forecasting, automated replenishment, and enhanced traceability. Drawing on the expertise of Ohta Vietnam, which has supported numerous implementations in the ASEAN region, including Vietnam, we will present concrete solutions.

First, let’s organize the specific challenges of screw inventory in the HMLV era and then delve into the solutions.

The Challenges of Increasingly Complex Screw Inventory in HMLV Production

The Background of the High-Mix Low-Volume (HMLV) Era

Due to a growing preference for customization and shorter market lifecycles, product lifecycles are shortening year by year. As a result, manufacturing sites have been forced to transition to a system of supplying “many product types” in “small quantities” with short lead times—in other words, High-Mix Low-Volume (HMLV). Fastening components like screws are directly linked to changes in finished product specifications, making them one of the parts most affected by the increase in variations. Parts procurement, which once simply involved bulk purchasing of general-purpose items, now requires sophisticated operations, from item number organization to replenishment timing.

The Swelling Management Load from Size, Material, and Lot Differences

• Diversification of Sizes: Part numbers are subdivided by head shape, length, and nominal screw diameter. It’s not uncommon for a single M3 screw to have dozens of variations in length and material.
• Combination of Materials and Surface Treatments: In addition to stainless steel and alloy steel, requirements for surface hardening and anti-rust coatings cause the number of stock-keeping units (SKUs) to increase exponentially.
• Differences Between Lots: In the automotive and medical device industries, lot tracking is mandatory, requiring separate storage and retrieval for each lot number, even for the same part number.

This complexity reduces inventory accuracy and makes it easy to create phantom inventory (inventory that exists on the books but is actually missing) and dead stock (excess inventory). Furthermore, assembly errors due to misidentified part numbers can lead directly to quality problems and recalls, meaning the management burden is more than just a cost issue.

Examples of Key Performance Indicators (KPIs)

KPI	Description	Guideline/Point
Inventory Turnover	Annual or monthly cost of goods sold ÷ Average inventory value	A rate of 5 times/year or more is ideal. Too high can risk stockouts.
Service Level	The rate at which customer demands were met immediately (inverse of stockout rate).	Aim to maintain 95% or higher while reducing inventory.
Holding Cost	The sum of warehouse rent, labor costs, and capital costs.	The target is 2-3% of sales. Can be compressed by organizing item types.

Inventory turnover indicates capital efficiency, while service level is a crucial indicator of availability that prevents line stoppages. These two are in a trade-off relationship, and finding a balance based on appropriate holding costs is the most important theme in screw inventory management in the HMLV era.

Visualizing Inventory with Classification Management and ABC Analysis

Suppressing Item Proliferation with Part Number Rules and Master Data Maintenance

If HMLV screw inventory is left as a disorganized “black box,” similarly shaped and sized parts get mixed up, leading to incorrect retrievals and inventory discrepancies. The first step is to establish a “part number system that anyone can uniquely identify” and “accurate master data.”

• Hierarchical Code Design
  • Example: M5-012-SUS-04 → Nominal Dia. – Length – Material – Surface Treatment
  • Linking drawing numbers and JIS/ISO standards with a suffix makes it resilient to design changes.
• Attribute-Specific Master Data
  • Maintain dimensions, material, and screw thread pitch in individual fields.
  • Link with IoT weight scales and WMS via API to automatically verify physical inventory against the master data.
• Regular Maintenance Rules
  • The purchasing department registers new part numbers, and obsolete ones are reviewed monthly.
  • Always ensure that shelf labels and system names match to prevent misidentification.

To accurately grasp the inventory quantity for each type, detailed classification and labeling are essential; the granularity of the classification determines the accuracy of management.

Determining Replenishment Priority with ABC + XYZ Analysis

Once part numbers are organized, the next step is to “visualize” them along the two axes of consumption value (ABC) and demand volatility (XYZ) to optimize replenishment logic.

	X: Low Volatility	Y: Seasonal Volatility	Z: High Volatility
A: High Value	AX (Kanban/Just-in-Time)	AY (Linked to demand calendar)	AZ (Safety stock + EPEI* shortening)
B: Medium Value	BX (Reorder Point)	BY (Leveling seasonal patterns)	BZ (Quote per order)
C: Low Value	CX (Periodic bulk order)	CY (Annual review)	CZ (Make-to-order)

*EPEI: Every Part Every Interval

Steps for Analysis

1. Calculate ABC Rank by Annual Usage ValueUsage Value Ratio = Annual Retrieval Quantity × Unit Price ÷ Total Value of All Parts
2. Calculate XYZ Rank by the Coefficient of Variation of Monthly (or Weekly) Consumption
   • Coefficient of Variation (CV) = Standard Deviation / Mean
   • CV ≤ 0.1: X, 0.1 < CV ≤ 0.25: Y, CV > 0.25: Z
3. Determine Operational Policy with the Matrix
   • AX: High risk of line stoppage, so respond immediately with IoT weight scales and automated replenishment.
   • CZ: If procurement cost is lower than holding cost, a zero-inventory operation is an option.

By accumulating and analyzing inventory data in the cloud, demand forecast accuracy improves, and reorder points can be adjusted automatically. At Ohta Vietnam, we have a track record of integrating this ABC/XYZ matrix into ERP systems and improving inventory turnover from an average of 120% to 160%.

Key Points

• Maintenance of part number and attribute master data is the prerequisite for “visualization.”
• Optimize replenishment rules for each part number using the ABC × XYZ matrix.
• Link with IoT weight scales and WMS to learn from consumption実績 in real time.
• Ranks are recalculated quarterly to track demand fluctuations.

This allows for preventing stockouts of high-value, high-volatility screws while suppressing excess inventory of low-value, low-volatility items, achieving both “optimal inventory × cost optimization.”

Achieving Real-Time Inventory Tracking with IoT Weight Scales and RFID

IoT Weight Scales for an 80% Reduction in Counting Man-Hours

Performing screw stocktaking and spot checks manually leads to a proportional increase in counting errors and labor load. By introducing IoT weight scales, load cells built into containers or shelves constantly monitor weight changes and calculate the exact quantity simply by “placing” the items, by multiplying by the mass per unit. Inventory information is automatically sent to the cloud, and an order alert is issued when a preset threshold is breached, minimizing the risk of stockouts.

In a case study from Ohta Vietnam, stocktaking that took 15 minutes per shelf with visual checks and counting scales was reduced to under 3 minutes, cutting counting man-hours by about 80%. Furthermore, by incorporating the measurement data into a weekly demand forecasting model, a system was built to dynamically review safety stock, compressing excess inventory by 15%.

Implementation Points

• Register the mass for each part number in the master data and set a quarterly calibration frequency.
• Choose PoE or long-life batteries for sensor power to minimize wiring costs.
• Link with WMS/ERP via API to automatically update reorder points and ABC ranks.

Achieving Zero Incorrect Retrievals with Handy Terminal Integration

After grasping inventory quantities with weight sensors, it is crucial to correctly link “people and things” using handy terminals that read RFID tags or barcodes. By scanning for a three-point match of the shipping kanban, the actual product tag, and the shipping instructions on-site, incorrect retrievals and picking errors can be prevented. Furthermore, if the system automatically writes to the WMS the moment an item passes through an RFID antenna, the discrepancy between system inventory and physical inventory is eliminated, achieving real-time inventory sharing.

At a client factory where Ohta Vietnam introduced RFID handy terminals, incorrect retrieval troubles, which had occurred at an average of 20 cases per month, were eliminated, reducing costs for re-inspection and emergency shipments by approximately 1.2 million JPY annually.

Operational Tips

• Attach RFID stickers to picking lists and scan the shelf number and part number simultaneously.
• Install fixed readers at retrieval gates to automatically supplement any missed handy terminal scans.
• Write lot numbers to tags to ensure traceability in subsequent processes.

By combining IoT weight scales and RFID, “automated quantity measurement” and “prevention of incorrect part/lot retrieval” can be achieved simultaneously, minimizing the risk of line stoppages even with HMLV screw inventory.

Optimizing Inventory Costs with Demand Forecasting and Automated Replenishment Logic

Guidelines for Choosing Between Reorder Point and Periodic Review Systems

• Reorder Point SystemThis is a method of “replenishing immediately when inventory falls below a threshold.” It has a high affinity with IoT weight scales and minimizes stockout risk. IoT scales can operate 24/7 and enable automatic ordering when a threshold is breached.
  • Suitable Cases:
    1. Parts with a stable lead time where the time to replenishment is predictable.
    2. Critical parts with a high stockout cost (leading to quality incidents or line stoppages).
    3. Environments where IoT or ERP can capture real-time inventory.
• Periodic Review SystemInventory is checked at regular intervals (e.g., every Friday), and the order quantity is calculated as “target inventory level – current inventory.” This is effective for reducing lot costs through bulk ordering for “special-order screws” with large monthly or weekly demand fluctuations.
  • Suitable Cases:
    1. Parts where demand fluctuations are predictable due to seasonal factors, but daily variations are large.
    2. When order lots are large and reducing order fees and transportation costs is a priority over frequent replenishment.
    3. When transportation efficiency is important, such as with regular deliveries from suppliers or container-based procurement.

Practical Tip: Hybrid management using the “reorder point system for stable items and the periodic review system for volatile items” can reduce the number of orders while also mitigating stockout risk.

Calculating Safety Stock with Time-Series Forecasting

1. Establish a Data Collection InfrastructureDaily and lot-specific consumption data acquired by IoT weight scales and handy terminals are automatically saved to the cloud and used as training data for demand forecasting models.
2. Model Selection and Demand Pattern Identification
   • Steady Base Demand: ARIMA models or exponential smoothing are effective.
   • Exogenous Factors like Campaigns/Equipment Load: Explain with regression analysis or Facebook Prophet.
   • Grouping of Similar Items: Utilize correlations with multivariate time series (VAR).
3. Safety Stock Calculation FormulaSafety Stock = Z × σLT
   • Z: Safety factor corresponding to the service level (e.g., 2.05 for 98%).
   • σLT: Standard deviation of demand during lead time.If the demand distribution during lead time can be estimated from a time-series model, σLT can also be dynamically updated, allowing for the maintenance of the target service level while suppressing excess inventory.
4. Implementation of Automated Replenishment Logic
   • Set up a workflow in the ERP/WMS triggered by “Forecasted Inventory < Reorder Point.”
   • Automatically generate purchase orders or EDI data and send them to suppliers, eliminating human subjectivity in ordering decisions.
   • As soon as delivery is recorded, feed the results back into the model to perform a rolling update of the forecast.
5. Verifying Effectiveness with KPIs
   • Inventory Turnover: Measure the improvement rate before and after implementing the replenishment logic.
   • Service Level: Calculated from the number of stockouts.
   • Total Holding Cost: Calculated from inventory value, warehouse costs, and stocktaking man-hours. There are cases where stocktaking man-hours were reduced by 80% with the introduction of IoT weight scales.

Summary (Practical Points)

• Visualize “when and how much to order” with numbers through real-time inventory × demand forecasting.
• The reorder point system + IoT weight scales is effective for preventing stockouts of critical parts.
• The periodic review system + group forecasting can suppress lot costs and freight charges.
• Directly link service level indicators to safety stock to set an inventory level that management can agree on.
• Link ERP/WMS with an automated replenishment system to eradicate human error and subjectivity.

Strengthening Lot Tracking and Quality Traceability

Shelf Arrangement to Enforce First-In, First-Out (FIFO)

Screws are prone to rust and lubricant degradation during long-term storage, so using the oldest lots first, FIFO (First-In, First-Out), is fundamental. The surest way to enforce this on-site is to design the storage space itself into a “FIFO-friendly” shape.

Measure	Specific Example	Effect
Gravity Flow Racks	Separate the loading and unloading sides and use roller shelves that push boxes forward.	Even in deep shelves, older items automatically move to the front, preventing mix-ups.
Color-Coding & Location Barcodes	Add lot numbers + colors to labels and link shelf number barcodes to the ERP.	The order of use is clear at a glance and can be read with a handy terminal.
Digital Signage	A monitor in front of the shelf lights up to display the “next lot to use.”	Workers can pick without hesitation, reducing training costs.

In practice, slogans like “Use the old ones first!” are difficult to enforce. By creating a mechanism where the shelf itself “pushes” the older lots forward, human error can be fundamentally suppressed.

Rapid Root Cause Analysis Flow for Defects and Recalls

In the unlikely event of a screw defect, being able to trace back “which lot was used in which product and when” within minutes is directly linked to minimizing line stoppages and recall costs.

1. Assigning a Unique Lot ID
   • Print lot IDs on screw bags, carts, and kanbans, or attach RFID tags.
   • Automatically register the manufacturing date and inspection results into the ERP upon scanning at receiving.
2. Inter-Process Tracing
   • Perform a two-point verification of the shelf barcode and lot ID with a handy terminal during picking.
   • On the assembly line, link screw-fastening machine data with the lot ID and store even the torque waveform.
3. Bidirectional Search of Production Output ⇄ Parts History
   • Instantly search for the screw lot used from the product serial number (downstream).
   • If a defective lot is found, extract a list of all affected products at once (upstream).
4. Root Cause Analysis and Correction
   • Conduct a 4M analysis (Man, Machine, Material, Method) to identify the cause of the defect rate exceeding the threshold.
   • If the cause is material-related, provide immediate feedback to the supplier and arrange for a replacement lot.

By combining RFID and barcodes, physical items, book records, and history match in real time, preventing quality incidents due to incorrect retrievals or mixing. In an implementation supported by Ohta Vietnam, the time to identify a recalled part was reduced from the conventional 8 hours to 15 minutes after strengthening traceability.

Key Points Summary

• Enforce FIFO not with “rules” but with the “structure of the shelves.”
• Manage lot IDs consistently with handy terminals or RFID and sync immediately with the ERP.
• Systematize bidirectional search (product → part / part → product) to instantly grasp the scope of a recall’s impact.
• In case of a defect, pursue the root cause with 4M analysis and provide feedback to suppliers and processes.

These measures enable the coexistence of quality traceability and inventory optimization even under HMLV conditions, significantly reducing the risk of line stoppages and customer complaints.

Ohta Vietnam’s Solutions for the Vietnamese Manufacturing Industry

Inventory Design Considering High Temperature and Humidity & Long-Distance Logistics

Vietnam has a high annual average humidity of around 78%, and iron/carbon steel screws are prone to rusting in this environment, making anti-rust and moisture-proof measures essential for both storage and transportation. Ohta Vietnam recommends the following inventory design.

Challenge	Solution	Effect
Rust/Oxidation from High Humidity	① Silica gel + VCI bag packing<br>② Use of corrosion-resistant grades like SUS410	Reduced rust incidence from an actual 0.8% to 0.1%.
Condensation from Temperature Differences	① “Cool zone warehouse” with 24-hour ventilation + dehumidifier<br>② Constant monitoring of temperature/humidity sensors with LoRaWAN	Real-time alerts for deviations in the storage environment.
Long-Distance/Multimodal Transport	① Vacuum packing + desiccant sealing before vanning<br>② Lead time correction assuming 5-14 days for sea transport	Eliminated in-container rust problems and maintained optimal inventory.

Furthermore, by adopting Vendor Managed Inventory (VMI), we directly monitor the customer’s warehouse inventory sensor values. By automatically shipping replenishment when the stock falls below the shortage threshold, we significantly reduce the user’s ordering tasks.

Success Story: 30% Improvement in Inventory Turnover / 60% Reduction in Stocktaking Time

Client: Japanese automotive supplier (Dong Nai Province)

Number of Items: 1,250 SKUs total for screws, washers, and spacers

Metric	Before	After	Improvement
Inventory Turnover	6.2 times/year	8.1 times/year	+30%
Stocktaking Time/person	15 h	6 h	▲60%
Stockout Incidents/month	4.8 incidents	0.7 incidents	▲85%

Keys to Success

1. Hybrid of IoT Weight Scales + RFID Gates
   • Acquired quantity data in seconds with weight scales and linked part numbers/lots with RFID.
   • Suppressed the error between system inventory and physical inventory to within 0.5%. This result is consistent with research reports that RFID contributes to inventory visualization and turnover improvement.
2. Demand Forecasting × Dynamic Safety Stock
   • Incorporated weekly retrieval data into a time-series model to automatically recalculate safety stock.
   • Reflected the variable lead times from long-distance logistics in the model’s parameters.
3. Standardization of Anti-Rust Packaging Specifications
   • Standardized the use of JIS Z 0310 VCI paper, creating a common specification for supplier-side packaging.
   • Reused the same VCI bags in front of shelves after unpacking to suppress operational costs.
4. VMI Contract + KPI Sharing Dashboard
   • Visualized turnover rate, service level, and holding costs in real time to align the interests of all parties.

As a result, we improved annual cash flow by approximately 230 million JPY while also shortening the lot tracking time during quality audits from 30 minutes to 5 minutes.

Key Points

• In Vietnam’s unique high-temperature, high-humidity environment, anti-rust packaging and temperature/humidity monitoring are essential.
• Achieve zero excess and zero shortages with dynamic safety stock calculation that considers long-distance lead times.
• By combining IoT weight scales, RFID, and VMI, both inventory accuracy and turnover rate can be improved simultaneously.

Ohta Vietnam will continue to expand its lean procurement & inventory optimization programs with an eye on collaboration between bases in the ASEAN region to support manufacturing in the HMLV era.

Conclusion

Key Takeaways

• In the era of High-Mix Low-Volume (HMLV), screw inventory increases exponentially due to variations in size, material, and lot, complicating management.
• Unify inventory with part number rules and master data maintenance, and clarify replenishment priorities with ABC × XYZ analysis.
• Combine IoT weight scales and RFID to sync quantity and lot information in real time, reducing stocktaking man-hours by 80%.
• Achieve both zero stockouts and improved inventory turnover through hybrid operation of reorder point and periodic review systems and dynamic safety stock based on time-series forecasting.
• Shorten the time to identify recalled parts from 8 hours to 15 minutes with shelf arrangements that enforce FIFO and consistent management of lot IDs.
• For Vietnam’s unique high-temperature, high-humidity, and long-distance logistics, anti-rust packaging, temperature/humidity monitoring, and VMI contracts are effective.

Checklist You Can Start on Tomorrow

1. Re-examine your part number system
   • Is it a code where nominal diameter, length, material, and surface treatment are clear at a glance?
2. Try calculating your inventory’s ABC rank
   • Tally up the annual usage value and try classifying into A (top 20%), B (next 30%), and C (remaining 50%).
3. Take a small step toward IoT adoption
   • Try a pilot introduction of a weight scale for just one item with “criticality A × demand volatility X” to visualize the effects.
4. Make simple improvements to your FIFO shelves
   • Consider whether you can “systematize” first-in, first-out with gravity racks or color-coded labels.
5. Start collecting demand data
   • Start recording daily retrieval quantities, even just in Excel, and use three months’ worth as seed data for a forecasting model.
6. Standardize the two-point verification of lot IDs
   • Enforce a rule to always scan the “shelf barcode + lot tag” during picking and leave a record.
7. Check your anti-rust measures immediately
   • Install VCI bags, desiccants, and temperature/humidity loggers in warehouse sections where humidity exceeds 70%.

By implementing these steps incrementally, you can steadily advance the optimization and cost reduction of your screw inventory.

Conclusion

For inquiries regarding samples or materials, please respond to this email or contact our sales representative!

At Ohta Vietnam, we not only provide ready-made products such as screws and bolts, but also support activities in the manufacturing field such as cutting, inspection, assembly, packaging…
Our equipment list is here

As a “specialized support” for the manufacturing industry in Asia. If you have any related questions, feel free to contact Ohta Vietnam for detailed support!
URL: https://ohtavn.com/en/contact/