Abstract: This paper presents a case study of the supply chain of a German cutting tool manufacturer, analyzed from a production logistics perspective by the Institute of Production Systems and Logistics (IFA). By applying the theory of logistic operating curves, the study identifies significant potentials for improvement in both manufacturing departments and product stores. The research highlights how separate optimization of supply chain tiers leads to poor global performance and proposes a collaborative Kanban-CONWIP control cycle to synchronize planning and achieve holistic supply chain objectives.
TL;DR
Local optimization within individual supply chain tiers often results in global inefficiency. This case study of a German cutting tool manufacturer demonstrates that applying logistic operating curves and transitioning to a collaborative, hybrid Kanban-CONWIP system can reduce throughput times by over 50% and eliminate significant inventory overshoots.
The Challenge: Bridging the Gap in SCM
Despite the strategic importance of Supply Chain Management (SCM), small to medium-sized enterprises (SMEs) often struggle to implement theoretical models. This study investigates a supply chain consisting of a cutting tool manufacturer and its primary supplier. Despite a trusting relationship and shared planning systems, the chain faced critical issues:
- Poor Availability: Inconsistent delivery to end customers.
- Bloated Inventories: Semi-finished stocks were 76% higher than necessary.
- Long Lead Times: Total replenishment time reached 60 shop calendar days (SCD), doubling the 30-day target.
- Conflicting Objectives: The supplier optimized for large lot sizes to minimize setup costs, which directly undermined the manufacturer’s agility.
Methodology: Logistic and Inventory Operating Curves
The IFA researchers employed two primary analytical frameworks to diagnose and address these inefficiencies:
- Bottleneck-Oriented Logistics Analysis: Utilizing throughput diagrams and logistic operating curves, this method identifies the work systems contributing most to total throughput time, rather than just focusing on machine speed.
- Logistics-Oriented Inventory Analysis: This approach uses inventory operating curves to model the mathematical relationship between mean inventory levels and service levels. It enables managers to identify the minimum stock required to maintain a target service level (e.g., 95%).
The study moves beyond local analysis by proposing a Collaborative Planning model. This involves replacing traditional push-based systems with a hybrid Kanban-CONWIP control cycle:
- Kanban manages "runners" with stable demand.
- CONWIP (Constant Work-in-Process) controls customized products, limiting backlogs by releasing orders based on actual bottleneck capacity.
Datasets
The analysis is grounded in real-world industrial data, including:
- Detailed order records (input/output dates and work content).
- Inventory levels for both semi-finished and final product stores.
- Customer demand forecasts versus actual delivery performance.
Key Insights: The WIP-Throughput Paradox
- WIP as a Lead Time Driver: High Work-in-Process (WIP) levels were the primary cause of long lead times. Reducing WIP by approximately 40% could improve speed without sacrificing output.
- The Cost of Local Optimization: The supplier's focus on large lot sizes resulted in a mere 5% delivery reliability, forcing the manufacturer to hold excessive safety stocks.
- Information Lag: Shared PPC (Production Planning and Control) systems are insufficient if information is passed with a time lag. True collaborative planning requires synchronized, transparent planning runs.
- Visual Trade-offs: Operating curves provide an objective tool for resolving partner conflicts by visualizing the inventory-service level trade-off.
Key Results
- Throughput Reduction: Fine-tuning WIP levels via operating curves can reduce process chain throughput time by more than 50%.
- Inventory Correction: The semi-finished store was identified as having a 76% inventory overshoot, acting as a warehouse rather than a Just-in-Time (JIT) transit point.
- Reliability Improvements: The supplier’s "invisible" 11-day backlog was exposed, explaining the 5% delivery reliability.
- Target Achievement: The proposed Kanban-CONWIP cycle makes the 30-SCD target lead time achievable for the first time.
Limitations and Future Work
While effective for a two-company model, real-world supply chains are complex networks. Future research aims to develop "process operating curves" to model entire supply chain networks as single holistic systems, further refining the resolution of conflicting objectives between partners.
Summary updated