Junzi Sun, Ph.D.

Funded research

Ongoing Tangram - open platform for modular, real-time, and data-driven aviation research
Tangram is an open research framework for flight surveillance data that originated as a hobby project for detecting turbulence. It has potential for various real-time aviation research topics such as GNSS jamming detection, aviation weather monitoring, emission analysis, and airport performance monitoring. The preliminary version is currently available as a web application developed in Python and JavaScript.
Funded by: NWO - Open Science Fund
Ongoing MOCHA Project - Multidisciplinary Open Collaborations on High Altitude Contrail Detection With Aviation and Satellite Data Fusion
This project is funded by TU Delft Climate Action Program. We are aiming at forming a small consortium with external industrial partners on remote sensing and flight data, and providing open-source tools, data, and machine learning models to support future research on the environmental impact of contrails.
Funded by: TU Delft - Climate Action Program Seed Fund
Ongoing NEEDED Project - Next generation data-driven reference European models and methods towards silent and green aircraft operations around airports
NEEDED responds to the second and third bullets of the “expected outcome” of the HORIZON-CL5-2022-D5-01-12 topic, delivering the next generation data-driven reference European models and methods to estimate present and future aircraft emissions.
Funded by: Horizon Europe
Ongoing Sustainable Aviation Lab
This lab is funded by the Aerospace Engineering Faculty at TU Delft. It aims to further understand and mitigate aviation’s environmental impact by harnessing different data sources with a hybrid of data-driven machine learning and model-based scientific approaches. As one of the two principal investigators, I lead the research areas in air traffic management, performance, data science, and machine learning.
Funded by: TU Delft Aerospace Engineering
2022 Open data and models to further Franco-Dutch collaboration on sustainable aviation
This project is funded by the Dutch Ministry of Foreign Affairs and Dutch Research Council under the Embassy Science Fellowship program. Between June and December 2022, I work with research partners in Toulouse on assessing the current environmental inefficiencies of aviation in Dutch and French airspace, in terms of excess emission indices. I will also study how future aviation policy would help further reduce the aviation emissions impact on society.
Funded by: NWO - Embassy Science Fellowship
2018 - 2022 Engage Knowledge Transfer Network
This project is funded by SESAR, H2020. Engage is managed by a consortium of academia and industry, with the support of the SESAR JU, to promote and facilitate the development of air traffic management research in Europe. The project focuses on inspiring new researchers and helping to align exploratory and industrial research, through a wide range of activities and financial support actions.
Funded by: H2020 - SESAR

PhD research

Ongoing Assessment and mitigation of global air transportation emissions

This research looks into data-driven approaches, making use of large-scale air traffic data and remote sensing data, to propose innovative solutions to accurately assess and reduce air transport’s environmental footprint in terms of emissions and contrails.

PhD Candidate: E.J. Roosenbrand
Ongoing Minimizing flight environmental impacts through operational optimizations

This PhD research looks into different optimization approaches to mitigating the environmental impacts of flights, with focus in at arrival and departure flights.

PhD Candidate: A. Tassanbi

Masters thesis research (ongoing)

Ongoing 3D wind field reconstruction using aircraft surveillance data

The research aims to use new machine learning models to generate 3D wind field from partially observable aircraft surveillance data.

Student: M. Slobbe
Ongoing Fair capacity balancing in the European airspace

The research explore methodologies for fairly managing and off-loading the capacities among sectors from the network manager perspective.

In collaboration with: EUROCONTROL Student: I Apahidean
Ongoing Efficacy of trajectory optimisation for contrail avoidance during hydrogen-powered flight operations

The research investigate the difference of hydrogen-powered aircraft in terms of contrail formation and how does it effect flight operations.

In collaboration with: NLR Student: O. de Koning
Ongoing Forecasting of Airline En-Route Delay for Individual Flights

This research uses supervised machine learning methods to predict en-route delay before departure from the airline perspective.

In collaboration with: KLM Student: C. Dolman
Ongoing Route optimization to avoid climate sensitive regions

This research aims to develop a flight route optimization strategy to avoid climate-sensitive regions.

Student: A. Greeve
Ongoing Large language model for air traffic control

This research explores the possibility of adopting large language models for air traffic control use cases.

Student: J. Andriuskevicius
Ongoing Modelling and investigating the feasibility and logistics of multi-modal transport

This thesis investigates the feasibility and environmental benefits of shifting short-to-medium range air passengers to ground transport, considering capacity and optimization for minimal impact and cost.

Student: B. Quadras

Masters thesis research (completed)

2024 Prediction of Aircraft Take-off Weight using Machine Learning

This research explores machine learning models to predicting the take-off weight based on flight plans and operation data.

Student: A.I. Gheorghe
2024 Feasibility and accuracy of Received Signal Strength-based Multilateration for aircraft localization using crowdsourced data

This study designs algorithms that performs multilateration using received signal strength. OpenSky data is used to evaluate the feasibility and accuracy of the proposed algorithms.

Student: V. Martjanova
2024 On Understanding Environmental Inefficiencies in Air Traffic Management: A Causal Inference Approach

The study assess the potential of causal analysis methods in uncovering the most significant inefficiencies in ATM operations.

In collaboration with: ANCE Student: J. Aalders
2023 Applying Large-Scale Weakly Supervised Automatic Speech Recognition to Air Traffic Control

This study explores fine-tuning the Whisper model for air traffic control, achieving significant error rate reductions, suggesting its potential for real-time application.

Student: J. van Doorn
2023 Dynamically Forecasting Airline Departure Delay Probability Distributions for Individual Flights using Supervised Learning

The research supports airline flight planning by developing machine learning models to forecast the departure delays, providing uncertainty and explainability.

In collaboration with: KLM & ATO Student: M. Beltman
2023 Machine learning based trajectory prediction to support demand forecasting

This research enhances air traffic sector demand forecasting, by employing a transformer neural network to generate accurate aircraft trajectories, improving stability in demand predictions.

Student: R. Vos
2023 Aircraft Tail-Specific Performance Modeling for Fuel Efficient Flight Operations

This thesis develops a tail-specific performance modeling framework using flight data and ML methods to optimize flight operations for fuel efficiency. It corrects biases in flight data, enabling post-flight analysis of fuel savings.

In collaboration with: KLM & ATO Student: F. Vossen
2023 Modeling and analyzing the environmental impact of short-to-medium range air and ground transports

Designing models to assess emissions and environmental impact of aircraft and ground transport modes.

Student: Y. Chen
2023 Automatic Speech Recognition for Air Traffic Control Using Open Data

This study aims at constructing an open RT corpus and an ATCo speech recognition model using neural networks and domain-specific information.

Student: J. Lubberding
2023 Streamlining multi-stop flights with ground transportation

Studying the inter-modal efficiency and analyze emission impact of layover flights in Europe using open data.

Student: K. Bislip
2022 Estimating Wind Fields Using Physically Inspired Neural Networks With Aircraft Surveillance Data

This research explores methodologies that incorporate dynamics of winds and improve the existing Meteo-Particle model under complex wind conditions.

Student: J. Malfiet
2022 Estimating wind fields using drones in a network

This research explores the estimation of wind for drones using low-cost airspeed sensors and existing onboard sensors.

Student: E. van Bassbank
2022 Trajectory optimization to minimize the environmental impact of departing and arriving aircraft

This research explores new ways to generate the most environmentally efficient trajectories. We also explore the environmental inefficiency in current standard arrival and departure practices and propose more efficient operational designs.

Student: L. van Dam
2022 Automatic Dependent Surveillance for Drones

This study looks into how an ADS-B compatible system for UAVs can be designed and implemented.

Student: S. Vlaskin
2022 Radio Frequency Fingerprinting for Aircraft Identification

In this research, we study how different surveillance signals for an aircraft can be processed and combined with machine learning approaches for the aircraft identification validation in CNS/ATM.

Student: A. Louwen
2021 Minimizing the environmental impact of cruise flights using meta-heuristic and optimal control optimizations

As part of the open aircraft performance (OpenAP) framework, in this research, we explored different optimization approaches to minimize the environmental impact of flight trajectories, with a focus for cruise flights.

Student: I. Govers
2021 Uncertainty Modelling in Aircraft Trajectory Predictions

This study investigated different methods to tackle the unpredictability of flight trajectories caused by various random factors. The goal was to model trajectory prediction uncertainty caused by these factors during different flight phases.

Student: R. Grass
2021 Long Short-Term Memory Network Based Trajectory Prediction Incorporating Air Traffic Dynamics

This study focused on the medium to long-term predictions of flight trajectories based on data-driven approaches. We aimed to model the dynamics of air traffic situations and incorporate this model to improve flight trajectory predictions.

Student: J.L. Overkamp
2021 ADS-B Signal Integrity and Security Verification Using a Coherent Software Defined Radio

This study developed a method to verify and validate ADS-B signal integrity with low-cost concurrent multi-channel software-defined radio receivers.

Student: W. Huygen
2021 Constructing aircraft emission model from open data

This study aimed at expanding OpenAP’s capability by developing an emission model, which can calculate emissions such as CO, NOx, CO2, and H2O based on ADS-B data.

Student: J. Jongbloed
2021 Modeling and Detecting Anomalous Safety Events Using ADS-B Data

The study proposed a proactive risk management strategy that uses safety indicators, which are obtained from ADS-B data using various data mining techniques. These safety indicators were then used to identify anomalous safety events and precursors at Schiphol Airport.

In collaboration with: ILT Student: A. Bonifazi
2018 ADS-B and Mode S data to enhance aviation meteorology and aircraft performance models

This study made use of open ADS-B and Mode S to construct accurate wind and temperature fields, and in turn, improved aircraft kinematic performance models.

Student: Q.H. Vû
2016 Aircraft performance parameter estimation using ADS-B data

This study made the first attempt to model flight envelope and estimate aircraft performance parameters using ADS-B data.

Student: T.W. Gloudemans
2016 ADS-B data and signal quality analysis for surveillance purposes

This research used MLAT data to assist the analysis of ADS-B data quality and to investigate the effects of internal and external factors affecting ADS-B implementation.

Student: T.L. Verbraak

© Copyright 2024 Junzi Sun