Optimizing the new product development process in building materials industry

Vicwest希望为新产品的引入开发一个可持续的流程模型，该模型基于数据，并将以最小的风险支持公司的业务计划。文献中的当前模型没有解决建筑材料行业特有的复杂性，例如竞争和市场的动态，以及缺乏综合行业数据。要设计这样一个过程，有几个挑战是建筑材料行业的特定性质所固有的，而不是广泛研究的零售商品部门。在这个项目中，我们专注于四个挑战：1）对该行业的新产品需求进行可靠的预测; 2）确定当前最佳的发布时间。 3）了解客户对新产品的接受模式; 4）将新产品需求预测与生产计划和库存管理相结合。我们将使用公司数据设计专门的扩散模型来预测新产品的需求。此外，我们将使用的需求预测作为输入的多阶段随机规划模型的生产计划和库存控制。我们还将探讨这个多阶段随机规划模型的解决方案。最后，我们将开发所提出的方法的原型。本研究将对学术界、产业界及HQP培训有所贡献。在学术界，我们将量化和分析不同的扩散模型与实际规模的行业数据;我们还将开发多阶段随机规划模型和相应的解决方法，用于新产品生产计划和库存控制。在工业领域，我们将帮助Vicwest开发模拟和优化决策支持工具的原型。在HQP培训中，我们将提供将业务分析专业知识应用于真实的行业应用的宝贵机会。

Vicwest would like to develop a sustainable process model for new product introduction that is based on data and will support the company's business plan with minimum risk. The current models in the literature do not address the complexities that are specific to the building materials industry such as the dynamics of competition and markets, and lack of aggregate industry data. To design such a process, there are several challenges that are inherent to the specific nature of the building materials industry, as opposed to the widely studied retail merchandising sector. In this project, we focus on four challenges: 1) developing reliable forecasts for demand of new products in this sector; 2) determining the optimal launch in regards to current company offerings as well as competing products; 3) understanding the customer adoption patterns of the new products; and 4) integrating the new product demand forecasting with the production planning and inventory management. We will use the company data to design specialized diffusion models to forecast the new product demand. Moreover, we will use the demand forecasts as inputs to a multi-stage stochastic programming model for production planning and inventory control. We will also explore the solution strategies for this multi-stage stochastic programming model. Finally, we will develop the prototypes of the proposed methods. This research project will make contributions to the academia, industry and HQP training. In academia, we will quantify and analyze the different diffusion models with real-size industry data; we will also develop multi-stage stochastic programming models and corresponding solution methodologies for new product production planning and inventory control. In industry, we will help Vicwest develop prototypes of the decision support tools of simulation and optimization. In HQP training, we will provide valuable opportunities in applying the business analytics expertise to a real industry application.