TL;DR
This paper proposes a novel, time-iterative approach to customer requirements (CRs) acquisition by integrating market-pull and technology-driven methods. This method aims to provide a more comprehensive understanding of past, current, and future customer needs, leading to more innovative product designs.
What Problem Does This Paper Address?
Acquiring accurate and comprehensive customer requirements (CRs) is crucial for product innovation and market success. However, existing methods often fall short. Market-pull methods, which rely on direct customer feedback, excel at capturing current needs but struggle with past or future-oriented requirements. Conversely, technology-driven methods, which analyze existing products and patents, are better at understanding past and predicting future needs but lack direct customer interaction, potentially leading to less accurate current requirement identification. This paper addresses the gap by proposing a unified approach that leverages the strengths of both methods across different time dimensions (past, present, and future) to achieve a more holistic understanding of CRs.
Methods
The core of the proposed method is a "time iteration" approach that categorizes CRs into three types:
- Past Customer Requirements (PCRs): Acquired through patent analysis and patent mapping technology. This involves analyzing patent technical information to identify the "effect" of a patent, which represents a PCR.
- Current Customer Requirements (CCRs): Identified using market-pull methods, including customer interviews and field investigations. The process involves extracting "jobs" from PCRs, building expert teams to formulate hypotheses about customer types, problems, and jobs, and then validating these hypotheses through customer interactions. CCRs are further classified into those evolved from PCRs and newly generated ones.
- Future Customer Requirements (FCRs): Predicted using evolutionary knowledge, particularly from the TRIZ framework and the Laws of Engineering System Evolution (LESE). This involves analyzing patent trends to predict future evolutionary states of a product.
These three types of CRs are then integrated to inform subsequent product design activities. The paper outlines a 7-step workflow for implementing this approach.
Datasets
The paper does not explicitly mention specific datasets used for training or evaluation in the traditional machine learning sense. Instead, it utilizes publicly available patent databases (e.g., from Zhihuiya.com) as the primary source of information for acquiring PCRs and predicting FCRs. Customer interviews and field investigations are also employed for identifying CCRs.
Key Insights and Ideas
- Time-based categorization of CRs is essential: Recognizing that CRs evolve and have different implications at different stages (past, present, future) is key to a comprehensive acquisition strategy.
- Synergy of market-pull and technology-driven methods: Neither method is sufficient on its own. Integrating them, particularly by using patent analysis to inform customer interviews and vice-versa, can overcome individual limitations.
- Patents as a rich source for historical and evolutionary insights: Patent analysis not only reveals past requirements but also provides clues about the trajectory of technological evolution, aiding in FCR prediction.
- Customer requirements are not static: The approach acknowledges that CCRs can be both evolved from past needs and entirely new, requiring distinct identification strategies.
- TRIZ and LESE offer powerful predictive capabilities: Evolutionary knowledge from these frameworks provides a structured way to forecast future needs, moving beyond subjective expert opinions.
Key Results
The paper presents a case study on hand installation pincers for spring band hose clamps. The application of the proposed method led to:
- Identification of PCRs such as "labor-saving" and "safety" from patent analysis.
- Determination of CCRs like "clamp repeatedly lead to labor intensity" and "too much process lead to low efficiency" through customer interviews.
- Prediction of FCRs such as "introduce a more efficient energy source" and "increase energy supply control for energy source" using evolutionary knowledge. The authors conclude that their integrated approach provides a more comprehensive and accurate way to capture CRs, leading to successful designs, especially for radical innovations.
Limitations and Future Work
- Chain-type evolutionary routes for FCR prediction: The paper acknowledges that FCR prediction currently only considers "chain-type" evolutionary routes. Future work could explore "tree-shaped" evolutionary patterns for a more comprehensive description.
- Uncertainty in radical innovation: The authors note that due to the uncertainty of radical innovation and the passive role of users, companies might need to rely more on technology-based and design-based innovation processes when customer contributions are limited.
- Further validation: While a case study is provided, more extensive validation across different product types and industries would strengthen the findings.
Summary updated