Mechanical product design per experiences of designers consumes unproportionally high manpower and time, as well as requires high relevance of designers’ experiences. To address this issue and extend the specific design innovation for general application, innovation methods of mechanical products have become a hot research topic. However, the currently available innovation methods are limited to specific needs of the market, exhibiting the form of discrete solutions. Thus, the innovation process is not sustainable. To this gap, this study integrates the emerging digital twin technology with engineering innovation theory and knowledge reuse concepts to establish a systematic sustainable innovation pathway (SIP) by inducing characteristic parameters. The pathway spares the demands for designer’s relevant experiences and avail sustainability of existing innovation methods. The derivation of the current SIP can be divided into six steps. To start with, the existing mechanical products based on users and market requirements are analyzed to synthesize general functional requirements. The characteristic parameters corresponding to the functional requirements are derived in the second step. Characteristic parameters are then specified and the potential problems which may occur in the induction of characteristic parameters are preliminarily solved. Digital twin technology is engaged to dynamically induce characteristic parameters. During the induction process, variation of related parameters due to changes in characteristic parameters are monitored. The problems that arise in variation may suggest new schemes, which solutions indicate new innovation directions. Finally, the candidate schemes of mechanical products according to different needs are identified. The formulated innovation plan and direction are archived in a knowledge base to establish a retrievable solution that matches the SIP. The applicability of current SIP is demonstrated by innovative design exemplification of a parking space management system.

机械产品设计每一个设计师的经验都会消耗不成比例的大量人力和时间，并且需要设计师经验的高度相关性。为了解决这个问题，并将特定的设计创新扩展到一般应用，机械产品的创新方法已成为一个热门的研究课题。然而，目前可用的创新方法仅限于市场的特定需求，呈现出离散解决方案的形式。因此，创新过程是不可持续的。针对这一差距，本研究将新兴的数字孪生技术与工程创新理论和知识重用概念相结合，通过引入特征参数建立系统的可持续创新路径（SIP）。该途径免除了对设计师相关经验的要求，并利用现有创新方法的可持续性。当前SIP的推导可以分为六个步骤。首先，根据用户和市场需求分析现有机械产品，综合通用功能需求。对应功能需求的特征参数在第二步中推导出来。然后指定特征参数，初步解决特征参数归纳中可能出现的问题。数字孪生技术用于动态诱导特征参数。在感应过程中，监测由于特征参数的变化而引起的相关参数的变化。变异中出现的问题可能会提出新的方案，这些解决方案表明了新的创新方向。最后，根据不同的需求确定机械产品的候选方案。制定的创新计划和方向在知识库中存档，建立与SIP相匹配的可检索解决方案。停车位管理系统的创新设计实例证明了当前 SIP 的适用性。