Edge AI for Predictive Maintenance in High Value Manufacturing

用于高价值制造中预测性维护的边缘人工智能

• 批准号: 10075464
• 金额: $ 6.28万
• 依托单位: AIRS ML LIMITED
• 依托单位国家: 英国
• 项目类别: Grant for R&D
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=10075464
• 关键词: Edge; AI; Predictive; Maintenance; Value

Machine breakdowns cause massive losses in production and efficiency. Predicting them is challenging.Traditional non-AI-based solutions may cause some/all of the following problems:1\. Unplanned downtime & Reduction in productivity2\. Severe accidents & Reduced lifetime of machines3\. Financial lossesAI has the potential to solve the aforementioned problems. However, current solutions rely on IoT/cloud computing, and suffer from the following drawbacks:1\. Power-hungry & Expensive2\. Data privacy issues3\. Increased carbon footprintTo quantify, a Cloud AI camera consumes 325 kg CO2 /year. This is because of the large amount of data transmission required for the camera.This project aims to create embedded system solutions that use edge AI to predict and prevent machine breakdowns.The solution analyses collected sensor data to monitor the machines and reduce the risk of a breakdown.The solution is:1\. Low-cost & Robust2\. Environment-friendly3\. Data secure (In-house)To quantify, an edge AI Camera consumes only 4 kg CO2 /year

机器故障会造成生产和效率的巨大损失。预测它们是具有挑战性的。传统的非基于人工智能的解决方案可能会导致以下部分/全部问题：1\.计划外停机和生产力降低2\。严重事故和机器寿命缩短3\。财务损失AI有可能解决上述问题。然而，当前的解决方案依赖于IoT/云计算，并且具有以下缺点：耗电量大且成本高2\。数据隐私问题3\。增加的碳足迹为了量化，Cloud AI摄像头每年消耗325公斤二氧化碳。这是因为摄像头需要大量的数据传输。该项目旨在创建嵌入式系统解决方案，使用边缘AI来预测和防止机器故障。该解决方案分析收集的传感器数据，以监控机器并降低故障风险。解决方案是：1\。低成本和坚固耐用2\.环境友好3。数据安全（内部）要量化，一台边缘AI摄像机每年仅消耗4千克二氧化碳