A mid-size precision parts manufacturer with 85 employees used targeted AI automation to slash defect rates, eliminate inventory mismatches, and generate RM480K in annual savings — all within a 6-week deployment window co-funded by MDEC.
0.8%
Defect Rate (down from 4.2%)
97%
Inventory Accuracy (up from 82%)
RM480K
Annual Savings Achieved
6 Weeks
Full Deployment Timeline
Precision Parts Manufacturing Sdn Bhd was losing ground to larger Tier-1 suppliers despite producing comparable quality components for the automotive and industrial sectors. Manual visual inspection on the production floor was yielding a 4.2% defect rate — enough for customers to raise concerns and threaten to dual-source. Compounding the problem, inventory records were misaligned with physical stock due to manual data entry, causing emergency purchases that eroded margins. Production reporting required overnight batch processing, meaning line managers were making decisions based on data that was 8 hours stale, with no visibility into real-time throughput or machine utilisation across the three-shift operation.
Deployed a computer vision quality inspection system (RM65K implementation cost, partially co-funded under the MDEC Smart Automation Grant) using industrial cameras and a custom-trained defect classification model at two critical inspection points on the line.
Built an AI-powered inventory forecasting module that ingested purchase orders, production schedules, and supplier lead times to generate weekly replenishment recommendations and flag discrepancy alerts automatically.
Implemented a real-time OEE (Overall Equipment Effectiveness) dashboard surfacing live throughput, downtime causes, and shift-level quality metrics to line managers and the factory director.
Conducted a 4-week structured change management programme to upskill floor supervisors on interpreting AI flags and feeding corrective actions back into the system loop.
Established a quarterly model retraining cadence tied to new defect samples, ensuring the vision system adapts as product variants and materials change.
Within three months of go-live, Precision Parts Manufacturing Sdn Bhd had demonstrably shifted from a reactive quality posture to a proactive one — a transformation that directly influenced a contract renewal decision from their largest automotive customer. The RM480K in annualised savings (from scrap reduction, emergency purchase elimination, and overtime avoidance) delivered a payback period of under five months on the total engagement cost, well within the threshold the CFO had set as a condition for board approval. More strategically, the deployment proved that a Bumiputera-owned SME with 85 employees can operationalise the same class of AI quality controls used by Tier-1 multinationals — without a dedicated data science team — by leveraging the right consulting partner and available government grant co-funding.
