Abstract
This paper takes the plastic waterproof junction box production line as the object of study. It analyzed enterprise production status, and then drawn the fuzzy value stream status map by using the value stream map and the combination of fuzzy triangular number method. It improved these problems existed in the fuzzy value stream status map, such as that inventory was too much because of push production, production cycle and changeover time was too long. It drawn the improved fuzzy value stream map. It makes delivery time shortened, and processing efficiency increased.
Introduction
In Lean Manufacturing, Value Stream Mapping (VSM) is the standard for visualizing material and information flow. However, traditional VSM relies on deterministic, "point-in-time" data, which fails to account for the inherent variability of real-world production—machine breakdowns, human variability, and supply chain fluctuations. This study addresses this "real-world noise" by analyzing "Workshop 3" at company "BM," which produces high-performance waterproof junction boxes. The workshop was plagued by typical "push production" symptoms: massive inventory piles, a sluggish 27.5-day production cycle, and long equipment changeover times.
Methodology
The researchers introduced a hybrid approach blending industrial engineering with fuzzy set theory to model this uncertainty:
- Fuzzy Triangular Numbers: Instead of deterministic values, the study used triplets $(l, m, u)$ representing minimum, most likely, and maximum possible times for each process. This provides a "buffer" in the model reflecting shop floor reality.
- Critical Path Analysis: By calculating fuzzy means and variances, the authors identified the sequence of stages dictating the overall lead time.
- Lean Optimization Tools: Once bottlenecks were identified, the authors applied ECRS principles (Eliminate, Combine, Rearrange, Simplify), transitioned to a U-shaped production line for better balancing, and implemented a Kanban-based pull system to eliminate inventory waste.
Key Results
The transition from the initial status map to the improved future map yielded dramatic improvements across all key performance indicators. While the actual value-creating time (the time spent physically shaping the product) remained nearly constant at approximately 14 minutes, the reduction in non-value-added time was transformative.
As shown in the data, the delivery cycle was slashed by 80%, from 27.5 days to just 5.5 days. This reduction in waiting time directly contributed to a 402.78% increase in working efficiency. Furthermore, the optimized layout allowed for a slight reduction in staffing from 16 to 15 people while maintaining higher throughput.
Conclusion and Future Work
This study demonstrates that combining VSM with fuzzy triangular numbers allows for a more accurate and actionable model of production environments. By switching to a pull system and optimizing the physical layout, the enterprise significantly reduced lead times and waste. Future research should focus on the scalability of this fuzzy model in high-mix, low-volume environments and the potential for real-time dynamic updates through integration with digital twin technologies.
Summary updated