Project "ARCA" - Automated Real-Time Classification in Aerospace Maintenance, Repair, Overhaul & Logistics market

航空航天维护、维修、大修和物流（MRO&L）在快速识别、解决和预防问题方面存在重大挑战和成本，这些问题会导致重大成本、延迟和最终安全。这是由于许多因素：系统复杂性的增加，不同问题的相互关联性以及分析数据的挑战。沃里克分析（WA）已被多家公司接洽，以解决航空航天工业中的无故障发现和其他经常性维护问题。WA在结构化数据方面取得了一些成功，但在脏数据、非结构化数据和异构数据方面遇到了挑战，这些数据不幸地成为行业的标准。MRO&L数据通常是手动收集的，要么是书面的，要么是通过计算机形式收集的。还有传感器“物联网”实时状态监测和故障诊断数据，这些数据通常以不同的格式收集。由于对问题或症状的不同解释和理解，书面数据通常在语义上不正确，有时在句法上也不正确（即错误）。这可能会导致难以找到问题的根本原因，从而难以解决问题或预测/防止将来的维护。最先进的技术是手动清理数据，重新解释和分类，以便可以使用。这是一个艰难的过程，最终会有缺陷。将其与传感器和诊断数据联系起来几乎是不可能的。这个原型开发项目建立在成功的概念验证项目的基础上，使用“真实世界”的非结构化、脏、异构航空航天数据创建一个新的行业解决方案航空航天根本原因分析（“ARCA”），以实时解决问题的根本原因，最终导致MRO&L中的成本效益解决方案，以实现有效、自动化、预测性和预防性维护，解决了行业中最大的总成本之一。

There are significant challenges and costs in aerospace Maintenance, Repair and Overhaul &Logistics (MRO&L) in terms of identifying, resolving and preventing issues quickly leadingto significant cost, delays and ultimately safety. This is due to a number of factors: increasingcomplexity of systems, interrelatedness of distinct issues and challenges with analysing thedata.Warwick Analytics (WA) has been approached by several companies to solve No Fault Foundand other recurrent maintenance issues in the aerospace industry. WA has had some successwith structured data, but has encountered challenges with the dirty, unstructured andheterogeneous data which are unfortunately the norm in the industry.MRO&L data are typically collected manually either written or through computer forms.There are also sensor “Internet of Things” live condition-monitoring and asynchronousdiagnostic data which are collected and typically in disparate formats. Written data are oftensemantically incorrect due to different interpretations and understanding of a problem orsymptom, and sometimes syntactically incorrect too (i.e. mistakes). This can make it difficultto get to the root cause of issues and therefore resolve them or predict/prevent maintenance inthe future. The state of the art is for data to be manually cleaned, reinterpreted and classifiedby hand so that it can then be used. This is a laborious and ultimately flawed process. Linkingit to sensor and diagnostic data is almost impossible by current means.This Development of Prototype project builds on the successful Proof of Concept project tocreate a new industry solution Aerospace Root Cause Analysis (“ARCA”) using ‘real-world’unstructured, dirty, heterogeneous aerospace data to resolve the root causes of issues in realtime,ultimately leading to cost effective solutions in MRO&L to enable effective, automated,predictive and preventative maintenance, tackling one of the largest aggregate costs in theindustry.