AI Process Mapper for the Waste & Recycling Industry

废物处理人工智能流程图

• 批准号: 10031051
• 金额: $ 22.19万
• 依托单位: SORTFLOW LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=10031051
• 关键词: AI; Process; Mapper; Waste; Recycling

SortFlow Limited is a London-based cleantech company which has developed, a platform to design, model and virtually test waste recycling processes.In this project, we will develop a new software module which takes data collected from optical sorting equipment and Artificial Intelligence (AI) vision systems in waste sorting and recycling plants, and integrates it into an AI powered live process map. The AI Process Mapper will provide to plant operators real time actionable insight into process performance, causes for underperformance and corrective action recommendations.As part of the project, we will process the data collected from pilot plants through machine learning models using 'supervised learning' to identify patterns and connections between the output qualities measured and sorting equipment infeed data in the sorting lines to understand the reasons behind performance variations and identify opportunities to improve, optimise processes and make performance predictions.The solution will optimise plant recycling processes and provide better recovery and purity of recyclable waste materials, which will in turn help to reduce the amount of recyclables being landfilled or incinerated (and the associated CO2 emissions) and increase revenue for waste and recycling facilities.

SortFlow Limited是一家总部位于伦敦的清洁技术公司，开发了一个设计、建模和虚拟测试废物回收流程的平台。在这个项目中，我们将开发一个新的软件模块，该模块将从废物分类和回收工厂的光学分类设备和人工智能（AI）视觉系统收集数据，并将其集成到人工智能驱动的实时流程图中。AI Process Mapper将为工厂操作员提供真实的可操作的流程绩效洞察，绩效不佳的原因和纠正措施建议。作为项目的一部分，我们将通过使用“监督学习”的机器学习模型处理从试点工厂收集的数据，以识别测量的输出质量与分拣线中分拣设备进料数据之间的模式和联系，以了解原因该解决方案将优化工厂回收流程，提高可回收废物的回收率和纯度，从而有助于减少填埋或焚烧的可回收物数量（以及相关的二氧化碳排放），并增加废物和回收设施的收入。