Turnaround prediction and controling with micrsocopic process modelling GMAN proof of concept & possiblities to use microscopic process szenarios as control options Bernd Oreschko, Thomas Kunze, Tobias Gerbothe, and Hartmut Fricke Chair of Air Transport Technology and Logistic Technische Universität Dresden, Germany {oreschko, kunze, gerbothe, fricke}@ifl.tu-dresden.de Abstract— For most flight phases automated and reliable target time predictions for an efficient resource management are common, but during the turnaround on ground best guessing by staff is still the standard. The turnaround prediction concept of TU-Dresden, called GMAN, is an approach to predict the Total Turnaround Time and the appropriate Target Off Block Time. The proof of concept in a real airport environment shows it ability to work reliable in an automated ATM-system, with suitable adjustments to the local information environement. Further an approach with microscopic process definition to offer control options is shown. Keywords: turnaround, predcition, A-CDM, processes I. MOTIVATION While for the airborne phase of flight, many approaches for a precise prediction of target times up to 30 seconds exist, the ground phase including the total turnaround time (TTT) is still out of the scope for reliable automated time prediction. The aircraft operators and manufacturers derive the TTT from deterministic sub-process (de-boarding, fueling, etc.) durations and their simple summation. This is not an accurate representation of reality as each of these sub-processes has a stochastic nature. This becomes especially evident in non- standard situations, e.g. delays or extended process duration. Nowadays ground staff solves possible conflicts using a best- guess behavior and not with the help of decision support tools to connect with the automated ATM-environment. [1][3][9][10][11] II. RESEARCH AIM AND REVIEW The scope of turnaround (TA) research at TU Dresden is to predict and control the turnaround of aircrafts and its processes by profound knowledge of all necessary activities and its dependencies. The core of the prediction is the so-called GMAN concept and application, based on several research activities - as further described - with the main idea of stochastic process description and stochastic time prediction in conjunction with microscopic process definition. Together with the main ideas of Airport Collaborative Decision Making (A-CDM) it can be used to optimize the use of airport infrastructure and capacity (as well as the individual processes themselves). A. A-CDM prediction gaps Within the A-CDM concept the main ideas are information sharing and the so-called milestone approach which establishes standard timestamps throughout every stage of ground operations, with the Target Off Block Time (TOBT) and Target Startup Approval Time (TSAT ) as most important timestamps. While the second one is reversely calculated from the Take Off Time (TOT) by well-known principals - for the first one no reliable calculating method is given. The most common way for issuing the TOBT is by operational staff’s best guess knowledge that, however, is not in accordance with the A- CDM goal for an always-reliable milestone definition. [5] [8] B. Previous Turnaround Research activities at TUD Fundamental knowledge for detailed process understanding was gathered during a study in cooperation with an aircraft manufacturer while aiming to understand turnaround reliability enhancements on a long time basis.[1] Significant contributions to uncertainty within the TA and potential for improvements in TA-reliability for future aircraft design showed up. One major impact factor contributing to uncertainties and non-standard process execution is the arrival delay, as shown in our study covering several German airports. It was also observed that airlines introduce dynamic scheduling buffers to mitigate the impact of disturbances in ground operation leading to TTT on their schedule integrity. [2] A following study focused on the processes themselves, analyzing effects on to the execution due to airport type. As shown in [3] beside the well-known airport categories hub and non-hub at least one additional classification can be found in the process execution – called supply basis. It was identified that the supply basis is a major reason for different observed process characteristics due to varying levels of staff skills based on different training principles and expertise [4]