Artificially Intelligent Monitoring System for Airport Baggage Handling Assets (AIMS)

处理不当的行李给乘客带来痛苦，对航空业来说代价高昂（2017年估计为每年17.6亿英镑），并损害了声誉。据估计，希思罗机场每1000个行李中就有1.6个因行李处理设备故障而处理不当。每天14万名乘客，每件行李约100英镑的赔偿相当于航空公司每天损失22400英镑。为此，希思罗机场制定了“每个乘客，每个行李，每次”的目标，行李系统停机时间少于半小时。不幸的是，大多数常见故障，通常是电机变速箱单元（MGU）或输送机，需要一个多小时才能修复。唯一的选择是在夜间停机期间进行维护。这通常也是不可能的，因为在设备故障严重到需要计划外停机之前，没有预先警告。复杂的状态监测系统已经存在，但它们只用于高资本价值的关键资产，因为它们过于昂贵而无法大规模部署（希思罗机场约30,000个mgu）。最好的系统之一是振动监测，因为它提供了大量关于关键机器部件（如齿轮、轴承和电动机）损坏的信息，以及诸如不对准、不平衡和裂纹等故障。预测振动监测所需的高质量加速度计成本约为500英镑，不适合在机场广泛部署。不幸的是，低成本的加速度计不能很好地检测早期损伤。电机电流特征分析（MCSA）是一种有效的、低成本的替代方法。目前，由于现有信号处理算法对电机电流数据的限制，其诊断能力也较差。AIMS将通过开发以下技术来克服这一挑战：a)使用低成本电流钳从电机电流数据中提取齿轮、输送机、轴承、电机和轴的早期损坏和故障的高保真诊断特征的新型信号处理。b)一种新型人工智能，能够对微控制器/输送机的健康状况进行分类。该系统的输出将用于为工程师创建易于解释的行李处理资产健康和优先维护行动的交通灯显示。该系统将广泛应用于机场和涉及快速消费品（FMCG）和仓储等机械处理的行业。AIMS将在项目结束后的5年内释放产生（贴现）2190万英镑毛利润的机会。26/04/21 -更新范围，将项目试验从希思罗机场改为史基浦机场

Mishandled bags create misery for passengers, are expensive for the airline industry (estimated at £1.76b p.a. in 2017), and damaging to reputations. An estimated 1.6 per 1000 bags are mishandled at Heathrow due to faults with the baggage handling equipment. 140k passengers per day and compensation of ~£100 per bag equates to airlines losing £22,400 a day. In response, Heathrow has set an objective of "Every passenger, every bag, every time" and baggage systems down-time of less than 1/2 hour.Unfortunately, most of the common failures, typically of a motor gearbox unit (MGU) or conveyor, take more than an hour to repair. The only option is to carry out maintenance during the nightly shut-down. This is not usually possible either, because there is no advance warning of equipment faults before they are serious enough to require un-planned shutdowns. Sophisticated condition monitoring systems exist, but they are only used for high capital value, critical assets because they are too expensive to deploy on a large scale (~30,000 MGUs at Heathrow).One of the best systems is vibration monitoring, because it provides a wealth of information on damage to critical machine parts such as gears, bearings and electric motors as well as faults such as misalignment, imbalance, and cracks. High quality accelerometers required for predictive vibration monitoring cost ~£500, which is untenable for widespread deployment at an airport. Unfortunately, low cost accelerometers do not perform well enough to detect early stage damage.Motor current signature analysis (MCSA) could be an effective, low cost alternative. Currently, it also suffers from poor diagnosis capability due to limitations of present signal processing algorithms for motor current data.AIMS will overcome this challenge by developing: a) Novel signal processing to extract high fidelity diagnostic features for early stage damage and faults to gears, conveyors, bearings, motors and shafts from motor current data using low cost current clamps. b) A novel artificial intelligence that is able to classify the health of MGUs/Conveyors. The output of this system will be used to create an easy to interpret traffic light display of baggage handling asset health and prioritised maintenance actions for the engineers. The system will have widespread application in airports and industries involving mechanical handling such as fast moving consumer goods (FMCG) and warehousing. AIMS will unlock the opportunity to generate (discounted) gross profits of £21.9m in a 5 year post project period.26/04/21 - update to scope to change project trials from Heathrow to Schipol airport