Research on Key Technologies of a Virtual Reality-based Product Test Platform

在产品开发的早期阶段，必须对各种客户需求和设计可行性进行有效的测试和评估，以获得产品最终满足客户满意的最佳机会。这需要对产品设计、制造和应用有深入的了解和广泛的信息。现有的产品测试和评估方法缺乏应对当前用户对产品复杂性、连接性和多样性的期望的能力。用户交互和产品模型没有完全集成到产品开发和应用程序环境中。本提案将开发一个用于产品评估和改进的虚拟测试平台。研究工作将集中在虚拟产品测试平台的关键技术上，目标如下：1)开发利用机器学习的智能方法，以减少产品评估和改进过程中决策对经验的依赖；2)开发基于手势的交互方法，使产品用户能够参与产品开发；3)虚拟测试平台的实现及应用于工业和消费产品的性能评价。拟议的研究将通过引入能够理解产品用户意图的过程来推进产品开发领域。这将导致一个自我进化的产品设计系统。本研究将朝着利用机器学习方法来总结设计知识并自动形成设计规则来指导独立于人类经验的产品开发的自动化设计迈出重要的一步。这是产品开发的新方向，是由不断增长的产品创新需求驱动的。建议的研究将为工业和产品用户开发工具，以在可控的计算机环境中测试和探索各种产品解决方案，为产品开发提供可行和切实的解决方案。它对下一代制造系统也至关重要，下一代制造系统在产品开发中采用物联网、数字制造能力、3D打印和大数据分析。

Various customer requirements and design feasibilities have to be effectively tested and evaluated in the early stage of product development to obtain the best chance for the product to eventually meet customer satisfaction. This requires deep knowledge and extensive information on the product design, manufacturing and application. Existing methods of product testing and evaluation lack the capability to cope with the current trend of user expectations for products that are characterized by complexity, connectivity, and diversity. User interactions and product models are not fully integrated in product development and application environments. This proposal will develop a virtual test platform for product evaluation and improvement. The research efforts will focus on key technologies of a virtual product test platform with the following goals: 1) development of intelligent methods using machine learning to reduce dependence of experience for decision-making during product evaluation and improvement; 2) development of a gesture-based interactive method to enable the involvement of product users in product development; and 3) implementation of the virtual test platform and evaluation of its performance by using industrial and consumer products. The proposed research will advance the field of product development by introducing processes that are capable of understanding the intent of product users. This will lead to a self-evolution product design system. This research will make a significant step towards automatic design using machine learning methods to summarize design knowledge and automatically form design rules to guide product development that is independent of human experience. This is a new direction in product development driven by the ever-growing need for product innovation. The proposed research will develop tools for industry and product users to test and explore various product solutions in a controllable computer environment that can provide feasible and tangible solutions for product development. It will also be essential for the next generation of manufacturing systems, which adopts the Internet of Things, Digital Manufacturing capabilities, 3D printing and Big Data analytics in product development.