Wan, J., Yar, K. P., Hean Low, M. Y., Loke, D., & Du, C. (2025). Optimizing CNC Spare-Parts Inventory with AI: Cutting Costs Without Compromising Availability. In (Editor), 2025 7th International Conference on Industrial Artificial Intelligence (IAI). https://doi.org/10.1109/iai68403.2025.11277188
Abstract:
Manufacturers of CNC spare parts face a
dual challenge: minimize holding costs while avoiding
stockouts of critical components. We propose an end-toend
AI pipeline that turns business records (customer
inquiries, sales transactions, inventory levels) into demand
forecasts and stocking recommendations. The pipeline
features: (i) automated data preprocessing (missing-value
imputation and outlier detection), (ii) sales-propensity
prediction via classification models (identifying which
parts are likely to sell), (iii) theoretical stock level
computation using Economic Order Quantity (EOQ) and
safety stock formulas, and (iv) customer segmentation
using Recency-Frequency-Monetary (RFM) analysis to
capture different purchasing patterns. All components feed
into a final module that integrates their outputs to produce
part-level stocking advice. We evaluate the method on an
industrial dataset of 9,909 CNC spare parts and 94,079
records (2015–2025). All identifying and cost information
was anonymized to preserve confidentiality without
degrading forecasting accuracy.
License type:
Publisher Copyright
Funding Info:
This research / project is supported by the A*STAR - RIE2025 Manufacturing, Trade and Connectivity (MTC) Industry Alignment Fund- Pre-Positioning (IAF-PP)
Grant Reference no. : M23L4a0001