Automated hazard detection for safe airport apron operations using a (weather-)robust LiDAR object recognition system

除了机场跑道是对航空旅行造成相当大风险的一个组成部分外，特别是机场机动性区域的特点是航空系统中安全事件的大量和经验证明的积累。由于邻近机场停机坪的操作人员、技术和程序标准化程度明显较低，迄今为止，机场降低风险的程序和技术措施仅集中在机动区域。在德累斯顿工业大学进行的初步研究表明，操作人员的信息情况存在缺陷，这将大大增加机场移动区域的风险。如果发生这种信息缺乏的情况，操作人员识别危险并随后避免损害的能力将明显受到限制。在这一点上，通过激光雷达传感实现的信息质量水平的提高被认为是有效减轻风险的关键因素。这一假设的相应证明将根据EUROCONTROL（一个申请人多年来积累了大量经验的领域）的既定风险评估程序的应用提供。然后，在方法学基础上确定风险缓解的潜力之后，将利用激光雷达传感器技术对停机坪上所有物体的3D点云进行实验验证，该技术将安装在德累斯顿国际机场。激光雷达被认为在可视化危险原因和与安全关键事件相关的衍生危险指标方面非常合格。更重要的是，激光雷达的这些基本特性产生了3D点云，因此收集的实验结果可以有条不紊地转移到其他传感器（如红外，雷达）。因此，将根据事故和事件数据对停机坪管制员的任务进行系统的危害评估，随后确定视觉指示器作为相应的原因。此外，将开发激光雷达监测概念，包括用于3D点云可视化的HMI概念，该概念将接受实验验证。与此同时，将在情景的帮助下对已确定的危险进行建模，并在运行HMI时进一步分析它们对减轻风险的估计贡献。在德累斯顿机场设想的实验现场测试范围内，将由运营停机坪管制员（由机场当局批准）重建和评估相关情景。将评价视觉手段在多大程度上有助于提高发现危害原因和有效减轻风险的可能性。德累斯顿机场的合作协议随附于本标书。

In addition to the airport runway being a component that incurs considerable risk to air travel, in particular airfield movement areas are characterized by a substantial and empirical proven accumulation of safety occurrences in the aviation system. Due to a significant low degree of standardization for operators, technology, and procedures on the neighboring airport apron, procedural and technological measures for risk mitigation at the airport have been exclusively focusing on the maneuvering area so far. Preliminary studies conducted at TU Dresden have shown deficiencies in the operators information picture that would significantly contribute to an increased risk on the airport`s movement areas. If such a lack of information occurs, the operator`s capability to recognize the hazard and subsequently avoid damage would be clearly limited. At this point, the augmentation of the information level of quality, enabled by LiDAR sensing, is considered as key element for effective risk mitigation. The respective proof of this hypothesis will be provided along the application of established risk assessment procedures according to EUROCONTROL, a field, in which the applicant has been gathering significant experience for years. Then, after the potential for risk mitigation will have been identified on a methodological basis, the respective potential will be experimentally verified by means of LiDAR sensor technology generating 3D point clouds of all objects on the apron which will be installed at Dresden International Airport. LiDAR is judged of being exceptionally qualified to visualize hazard causes and derived hazard indicators connected to safety critical occurrences. Even more, these fundamental characteristics of LiDAR generated 3D point clouds and thus so gathered experimental results can methodically be transferred to other sensors (e.g. infrared, RADAR).As such, a methodical hazard assessment of the apron controller`s task based on accident and incident data will be conducted, subsequently followed by the identification of visual indicators as corresponding cause. Further, a LiDAR monitoring concept will be developed, including an HMI concept for a 3D point cloud visualization that will be subject to experimental validation. Along this, identified hazards will be modeled with the help of scenarios and will be further analyzed on their estimated contribution to risk mitigation while operating the HMI. In the scope of an envisaged experimental field-test at Dresden Airport, relevant scenarios will be reconstructed and assessed by operational apron controllers (granted by the airport authority). It will be evaluated how far a contribution to a higher detection probability of hazards causes and effective risk mitigation can be achieved by visual means. The respective cooperation agreement from Dresden Airport is attached to this tender.