Abstract
In the era of mass customization (MC), manufacturers must balance the demand for high product variety with the need for production efficiency. This paper addresses the limitations of static product platforms by proposing a dynamic planning model for platform and module identification. Utilizing graph theory, the research analyzes the versatility and physical consolidation of product variants through directed and undirected graphs. A dynamic planning algorithm is introduced, allowing platform thresholds and module boundaries to be adjusted based on market demands. The effectiveness of this approach is demonstrated through a case study involving the component planning of various humidifier models.
Introduction: The Mass Customization Paradox
Mass Customization (MC) presents a fundamental manufacturing challenge: providing diverse, customized products at costs comparable to mass production. The traditional solution is "platforming"—the creation of a common core of components shared across a product family. However, most existing platforming strategies are static. Once a module is defined, it remains fixed, making it difficult for manufacturers to adapt to rapid shifts in consumer preferences.
This research proposes a shift from static to dynamic planning. By treating product components as nodes in a graph and adjusting platform boundaries in real-time, manufacturers can evolve their product designs to better align with current market trends without necessitating a total redesign of the production line.
Methodology: A Graph-Theoretic Approach
The core of the proposed framework lies in its use of Graph Theory to model the complex relationships between product components. The authors employ a dual-graph representation to capture different dimensions of product architecture:
- Directed Graphs (Versatility): These are used to express the "flow" and versatility of parts across different product variants. They help identify which components are most frequently utilized across the entire product family.
- Undirected Graphs (Physical Consolidation): These model the physical and functional integration within an assembly. They represent how parts are connected, ensuring that identified modules are physically coherent and manufacturable.
The Dynamic Planning Algorithm
The innovation of this model is an algorithm that avoids fixed boundaries. Instead, it utilizes "platform thresholds" and "module boundary parameters" that can be tuned dynamically. If market demand for a specific feature increases, the algorithm can redefine which parts belong to the "standard" platform versus the "custom" modules.
Furthermore, the model includes a feature for Potential Part Extraction. This identifies components that currently sit outside the platform but demonstrate high versatility across variants, suggesting they are candidates for future standardization.
Case Study: Humidifier Product Families
The authors validated their model using a practical engineering dataset of humidifier product families. By analyzing the components and assembly structures of various humidifier models, the algorithm was able to:
- Identify core platform components that should remain standardized.
- Group auxiliary parts into logical, swappable modules based on demand.
- Proactively suggest parts for future standardization based on emerging usage patterns.
While the study confirms the qualitative effectiveness of the approach in organizing complex part hierarchies, it also highlights the necessity of robust market data to feed the algorithm's threshold parameters.
Insights and Future Directions
The research suggests that a product platform should be viewed as a "living" entity rather than a fixed blueprint. By optimizing for both versatility (how often a part is used) and consolidation (how well parts fit together), manufacturers can achieve a more responsive production cycle.
Limitations and Future Work
- Scalability: While effective for humidifiers, applying this graph-based approach to highly complex products like automobiles—which contain thousands of parts—may require significant computational optimization.
- Data Integration: The mechanism for translating raw market data into specific graph thresholds remains a complex challenge.
- AI Integration: Future research may explore the use of machine learning to automate the adjustment of platform thresholds in real-time as sales and trend data are ingested.
Conclusion
By leveraging graph theory, this dynamic planning model provides a mathematical foundation for flexible manufacturing. It allows designers to see "around the corner," identifying parts that are becoming de facto standards and adjusting the product architecture to maintain a competitive edge in a fluctuating market.
Paper Citation: Wu, Y., Zhang, H., Luo, L., & Xu, Y. (2017). Dynamic Planning for Product Platform and Module Based on Graph Theory. Manufacturing Technology, 17(5), 70-71.
Summary updated