PhDs receiving financial support through Engage.
Decision support system for airline operation control hub centre (‘DiSpAtCH’)
The general objective of the DiSpAtCH (Decision Support System for Airline Operation Control Hub Centre) project is to elaborate on artificial intelligence technologies and how these technologies could efficiently support decision making in an Airline Operation Control Hub Centre (OCC) in unexpected or very complex situations.
The DiSpAtCH project aims to develop a novel approach for an airline operation control centre decision support system. Many different sources of information may influence the actions of persons working in airline operation control centres. Obviously the most challenging task is the information management task. This task includes monitoring, recognition and projection of relevant information out of all information available including current and future situations.
With the use of machine learning algorithms like supervised algorithms, abnormal multivariate situations can be detected or information classification can be done. Furthermore unsupervised algorithms support the knowledge discovery of undefined dependencies in the diverse information available. It will improve, for example, the description of any disruption within an OCC.
With a decision support system like DiSpAtCH the decisions within an airline operation control centre can be optimized which will result in fewer delays for passengers and better utilization of resources on airline level.
Trajectory planning for conflict-free trajectories: a multi agent reinforcement learning approach (RL4CFTP)
The objective of this PhD is to explore and present novel algorithms towards planning conflict-free trajectories at the pre-tactical phase of operations in computationally efficient ways, for a large number of trajectories in airspaces comprising multiple FIRs, following a methodology combining data-driven and agent-based approaches.
While data-driven methods aim to build models for trajectory planning and conflicts resolution incorporating stakeholders’ interests and preferences, the multi-agent reinforcement learning (MARL) approach aims to address complexity phenomena due to traffic, and resolve conflicts between multiple trajectories. Towards that goal we aim to formulate the problem as a Markov Decision Making Process and apply multi-agent reinforcement learning (MARL) methods to resolve it.
Detection, classification, identification and mitigation of GNSS signal degradations by means of machine learning
(abstract to follow)
Machine learning techniques for seamless traffic demand prediction
The overall goal of the proposed PhD study is to investigate innovative approaches to air traffic demand forecasting based on the synergistic combination of physical models of aircraft trajectories with artificial intelligence and machine learning techniques, with particular focus on pre-tactical traffic forecast, when most flight plans are not yet available. The primary approach will be based on the aggregation of the trajectories estimated by means of a set of hybrid models combining data-driven and model-based methods, but the PhD will also explore the direct estimation of traffic volumes and complexity by means of machine learning techniques. The study, which will be conducted in collaboration with the Network Manager, is expected to contribute to the provision of improved demand forecasts in a TBO environment and to support the definition of advanced ATFCM services.
Machine learning applications to extend AGENT’s conflict resolution capabilities
Air traffic management’s (ATM) aim is to ensure separation management of aircraft in an efficient way, minimizing possible delays and costs. The continuous increase in air traffic demand requires a higher level of automation to support the decision making. Adaptive self-Governed aerial Ecosystem by Negotiated Traffic (AGENT) was an exploratory research project supported by the H2020 Research and Innovation Program. AGENT proposes a system where the avoidance of potential loss of separations is done in a distributed, collaborative way between the involved aircraft while the controller is monitoring the process, preserving situational awareness. This proposal, built on AGENT’s future work proposals, seeks to investigate possible improvements of several critical aspects of the system through applications of various Machine Learning techniques. Specifically it seeks to improve the accuracy of the prediction of the deadlock (the hard deadline to reach an agreement through negotiations), attempts the construction of a possible “ecosystem complexity” estimator and investigate the effect that the introduction of agents’ learning capabilities can have at the negotiation dynamics. The work of the PhD candidate is being held at Autonomous University of Barcelona (UAB) in collaboration with ASLOGIC and under the supervision of Dr. Miquel Angel Piera.
Integrating weather prediction models into ATM planning (‘IWA’)
Weather has a big impact on air traffic management (ATM). Inefficient weather avoidance procedures and inaccurate prognosis lead to longer aircraft routes and, as a result, to fuel waste and increased negative environmental impact. A better integration of weather information into the operational ATM-system will ultimately improve the overall air traffic safety and efficiency.
The Air Transportation research group within the Communications and Transport Systems (KTS) at Linköping University (LiU), together with the Research and Development at Luftfartsverket (LFV, Swedish ANSP) develops optimization techniques to support efficient decision-making for aviation authorities. In a series of projects we plan to refine our frameworks, making them more robust with respect to changing weather conditions influence. The student will study and apply probabilistic models and the corresponding weather data, and to prototype and test the mathematical tools which will help to account for influence of bad weather conditions on the solutions developed in our related projects.
We work in collaboration with University of Sevilla, Spain, Professors Damian Rivas and Antonio Franco.
Advanced statistical signal processing for next generation trajectory prediction
It is expected that this PhD will contribute with better and more robust algorithms for trajectory prediction tools for the execution phase of the flight, which could support new or enhanced tools for advanced air traffic services into a Trajectory Based Operations (TBO) environment, increasing in this way, the safety capacity, predictability and cost-efficiency of the future European ATM system. This PhD will look at the trajectory prediction problem from a new probabilistic perspective, approaching it with powerful mathematical tools arising from the statistical signal processing field. Such advanced robust statistical inference techniques have been shown in other contexts to provide a remarkable performance/robustness improvement with respect to conventional approaches, then probably leading to significant contributions in the trajectory prediction field.
A pilot/dispatcher support tool based on the enhanced provision of thunderstorm forecasts considering its inherent uncertainty (‘STORMY’)
Thunderstorms represent a major source of disruption, delays and safety hazards in the ATM system. They are challenging to forecast and evolve in relatively rapid timescales. Even for the most advanced met products (which most stakeholders lack of), thunderstorm forecasts are provided in a deterministic manner. Both met provision and ATM use of met information need to consider the uncertainty in the forecasted evolution of these phenomena.
This PhD pursues three scientific goals, namely:
The PhD thesis would require a multidisciplinary approach, including disciplines such as meteorology, statistics, control, and ATM. We have built a consortium to build up multidisciplinary research, yet incorporating industrial actors (dispatchers) and stakeholders (met offices).
We will use different ground breaking methodologies to approach the different problems: on the one hand, advanced statistical methods (e.g., Bayesian model averaging, krieging) for the calibration, probabilistic fitting, and assimilation of thunderstorms forecasts and data; then, different strategies for short-term trajectory planning under uncertainty, which combine stochastic optimal control techniques, differential algebra, and GPU based.
Second generation agent-based modelling for improving APOC operations
(abstract to follow)
Stochastic control of tactical airline operations in hub airport networks
This dissertation project aims at developing a decision support system for Airline Ground Operations Controllers at Hub Airports. Focal point of the research is the introduction and evaluation of standardized schedule recovery actions which consider the situational state of network operations at a tactical time horizon. For enabling a proactive cost-impact-assessment of potential recovery actions, the concept foresees a microscopic optimization model for stochastic ground operations and seeks to integrate sub-optimal solutions and constraints from inbound, aircraft turnaround and downstream network operations. The model output is analyzed for selected disruption scenarios according to different airline-internal decision paradigms. Additionally, the competition among airlines for specific airport resources is investigated with various market mechanisms in order to explore how new courses of action for airline controllers might expand the solution space and benefit the ATM system performance.