SCORE: Supply Chain Optimisation for demand Response Efficiency

这个工业研究项目的愿景是从TRL3到TRL5， SCORE：需求响应效率的供应链优化。该系统将使制造业的一级和二级供应商能够通过数字技术更好地管理库存，并最大限度地减少需求和维护活动突然变化的影响。实现这一愿景的关键目标是：1。确保物料在供应链各节点之间顺畅流动。通过跟踪不同阶段的物料，尽量减少生产的等待时间，避免延误。根据生产线的生产单元周期自动化原材料需求，以尽量减少“在地板上”未使用的材料4。集成持续学习模型，用于预测需求和机器故障本项目的主要重点是实现用于跟踪和跟踪库存的传感器，以及开发用于创建需求预测模型和库存变化模型的机器学习算法。虽然企业资源计划（ERP）系统考虑了一些因素，例如计划的维护活动，但它们大多是通用工具，缺乏专业的预测系统，并且广泛依赖统计方法进行库存控制预测。SCORE的创新之处在于应用机器学习来优化传统上使用统计分析方法的供应链管理模型，将不同的模型集成到一个模型中，并与整个供应链进行预测沟通，从而更精确地控制库存，提高资产的可追溯性，并几乎消除供应延迟或零件积压。我们最初的目标市场是供应链管理（SCM）软件市场，以一级和二级供应商为目标用户。该项目代表了英国供应链管理的一个明确的技术创新，以及中小企业供应链联盟的主要增长机会。为了成功实现这一目标，项目联盟具有相关的专业知识，包括跟踪和跟踪系统开发、机器学习算法开发和库存控制专业知识。

The vision of this industrial research project is to bring from TRL3 to TRL5, SCORE: Supply Chain Optimisation for demand Response Efficiency. This system will enable Tier 1 and Tier 2 suppliers in manufacturing sectors to better manage their inventory through digital technologies and minimise the impact of sudden changes in demand and maintenance activities.The key objectives in fulfilling this vision are to:1. Ensure smooth flow of materials between different nodes of supply chain2. Minimise waiting time to start production and avoid delays through tracking of materials at different stages3. Automate raw material demand according to production cell cycles for production lines to minimise 'on-floor' unused material4. Integrate continuous a learning-enabled model for prediction of demands and machinery breakdownsThe main areas of focus in this project are on implementing the sensors for the track and trace of inventory and developing machine learning algorithms for the creation of demand forecast model and inventory change models. Although enterprise resource planning (ERP) systems take into consideration some factors, e.g. the scheduled maintenance activities, they are mostly generic tools, lacking specialist forecasting systems, and relying extensively on statistical methods for inventory control predictions.The innovation in SCORE lies in the application of machine learning to optimise supply chain management models which traditionally use statistical analysis methods, the integration of different models into one and the communication of the forecasts with the entire supply chain, leading to more precise control over the inventory, greater traceability of assets, and near elimination of delays in supply or overstocking of parts.Our initial target market is the supply chain management (SCM) software market, with Tier 1 and Tier 2 suppliers the target users. This project represents a clear technological innovation for UK SCM, and major growth opportunity for the SME supply chain consortium. To successfully achieve this, the project consortium features the relevant expertise including track and trace system development, machine learning algorithm development, and inventory control expertise.