Theses
Internships / research assistantships
Our department does not offer any such positions for external BSc or MSc students.
BSc projects
See BrightSpace for Bachelor End Projects for TU Delft students. Only in rare exceptions, we can accommodate external students (and they definitely need to be self-funded, e.g. via Erasmus+).
MSc projects
See BrightSpace for Bachelor End Projects for TU Delft students. Only in rare exceptions, we can accommodate external students (and they definitely need to be self-funded, e.g. via Erasmus+).
BSc projects
Vacancies for projects again from Q2 22/23 onwards. I will be able to accept a few TU Delft students of the MSc Robotics, of the Vehicle Engineering track and BioRobotics as well as Haptics Interfaces specialization within the MSc Mechanical Engineering (see BrightSpace for project descriptions). For all other TU Delft MSc programs, tracks, and specializations see below. Only in rare exceptions, we can accommodate external students (and they definitely need to be self-funded, e.g. via Erasmus+).
Initiative Applications
We can occasionally accommodate (additional) highly motivated students in MSc projects or in ongoing projects before their graduation stage.
If you are interested in working on a project with us, please send us ca. 1 academic quarter before the intended starting date an e-mail including:
* a short CV (½ - 1 page) * BSc & MSc transcripts * a short motivation letter * BSc thesis (PDF) * any other information that might be relevant for the applicationThe motivation letter should state
* what you are interested in (control/ML topic and/or application) * the type of assignment you are interested in (more theoretical/more applied) * whether you prefer a project at the TU or at a company * the intended starting date * your relevant experience (studies, technical projects, internships, hobbies, etc.) * programming languages and related (Python, C, C++, ROS, etc.)Ongoing Theses
- Zheyu Du
Mentor(s): Giovanni Franzese (TU Delft), Jens Kober (TU Delft)
"Interactive Reinforcement Learning for In-Hand Object Manipulation" - Yurui Du
Mentor(s): Son Tong (Siemens Digital Industries Software), Flavia Sofia Acerbo (Siemens Digital Industries Software), Jens Kober (TU Delft)
"Autonomous vehicle learning control via multi-agent driving simulator" - Jianyong Sun
Mentor(s): Jihong Zhu (TU Delft), Jens Kober (TU Delft)
"Learning interactive tasks in a dynamic environment" - Yulei Qiu
Mentor(s): Jihong Zhu (TU Delft), Jens Kober (TU Delft)
"Learning robotic cloth flattening" - Alexander Keijzer
Mentor(s): Bas van der Heijden (TU Delft), Jens Kober (TU Delft)
"Frequency Domain Analysis for Experience Selection in Deep Reinforcement Learning" - Annemarijn Haasdijk
Mentor(s): Sergio Grammatico (TU Delft), Tim Biemans (Picnic), Jens Kober (TU Delft)
"The Quasi-Online algorithm in a Robot Packing environment" - Rudy De-Xin de Lange
Mentor(s): Dimitris Karageorgos (HIT), Jens Kober (TU Delft)
"Autonomous manipulation tasks and learning in a health-care environment" - Gijs van den Brandt
Mentor(s): Jari van Steen (TU/e), Alessandro Saccon (TU/e), Jens Kober (TU Delft)
"Learning From Demonstration for Manipulation On A Dual-Arm Setup" - Bram Bornhijm
Mentor(s): Tom van Riet (AMC), Jens Kober (TU Delft)
"Understanding tooth removal through motion primitive learning" - Besim Bay
Mentor(s): Ravi Prakash (TU Delft), Jens Kober (TU Delft)
"Learning object-object affordances" - Sven Uitendaal
Mentor(s): Alessandro Saccon (TU/e), Jens Kober (TU Delft)
"Impact-Aware Learning from Demonstration"
Completed Theses
- Dirk-Jan Boonstra
"Towards Corrective Deep Imitation Learning in Data Intensive Environments - Helping robots to learn faster by leveraging human Knowledge."
Mentor(s): Rodrigo Perez Dattari (TU Delft), Carlos Celemin(TU Delft), Jens Kober(TU Delft) - Irene Lopez Bosque
"Towards Corrective Deep Imitation Learning in Data Intensive Environments - Helping robots to learn faster by leveraging human knowledge"
Mentor(s): Rodrigo Perez Dattari (TU Delft), Carlos Celemin (TU Delft), Jens Kober (TU Delft) - Armin Avaei
"Learning Human Preferences for Motion Planning in Robotic Manipulation"
Mentor(s): Linda van der Spaa (TU Delft), Jens Kober (TU Delft), Luka Peternel (TU Delft) - Karthik Arvind Arunmoli
"Data efficient pick location learning from images for sequential tasks"
Mentor(s): Rodrigo Perez Dattari (TU Delft), Jens Kober (TU Delft) - David ter Kuile
"Handling the unknown: Towards on-the-job object recognition"
Mentor(s): Osama Mazhar (TU Delft), Jens Kober (TU Delft) - Chia-Yu Tsai
"Wrinkle contraction direction: a useful feature for learning robotic fabric manipulation from demonstration"
Mentor(s): Jihong Zhu (TU Delft), Jens Kober (TU Delft) - Lalith Keerthan Suresh Kumar
"Learning Task Space Policies from Demonstration"
Mentor(s): Carlos Celemin (TU Delft), Jens Kober (TU Delft) - Bram Harleman
"Automated modelling of traffic signal controller behaviour from floating car data using Hidden semi-Markov Models: Paving the way for large scale implementation of Green Light Optimal Speed Advisory systems"
Mentor(s): Andreas Hegyi (TU Delft), Jens Kober (TU Delft) - Evelyn D'Elia
"Automatic Controller Selection on a Humanoid Robot: Using Optimization Techniques"
Mentor(s): Serena Ivaldi (INRIA), Jean-Baptiste Mouret (INRIA), Jens Kober (TU Delft) - Yen-Lin Wu
"One Pose Fits All: A novel kinematic approach to 3D human pose estimation"
Mentor(s): Osama Mazhar (TU Delft), Jens Kober (TU Delft) - Anna Mészáros
"Teaching Robots How to Grasp Like Humans by Humans: An Interactive Approach"
Mentor(s): Giovanni Franzese (TU Delft), Jens Kober (TU Delft) - Niels Lander
"Interactive Imitation Learning for Force Control"
Mentor(s): Carlos Celemin (TU Delft), Jens Kober (TU Delft) - Bart Bootsma
"Learning Interactively to Resolve Ambiguity in Sensor Policy Fusion for Robot"
Mentor(s): Giovanni Franzese (TU Delft), Réka Bérci-Hajnovics (DoBots), Jens Kober (TU Delft) - Nathan Timmers
"Simulating Gait with the 3R60 Knee Prosthesis and a Control Moment Gyroscope"
Mentor(s): Saher Jabeen (TU Delft), Heike Vallery (TU Delft), Jens Kober (TU Delft) - Arthur Helsloot
"Symbolic Reasoning about Unseen Objects from Multimodal Sensory Feedback for Manipulation"
Mentor(s): Mert Imre (TU Delft), Carlos Hernandez Corbato (TU Delft), Jens Kober (TU Delft) - Willem de Graaf
"Understanding Tooth Removal Procedures using Feature Engineering and Classification Models"
Mentor(s): Jens Kober (TU Delft), Tom van Riet (AMC) - Peter Valletta
"Uncertainty-based Interactive Machine Learning"
Mentor(s): Rodrigo Perez Dattari (TU Delft), Jens Kober (TU Delft) - Olav Jacobs
"Cooperative Robotic Arms for Parcel Picking"
Mentor(s): Jan-Willem Damsteeg (PrimeVision), Jens Kober (TU Delft) - Snehal Jauhri
"Interactive Learning in State-space: Enabling robots to learn from non-expert humans"
Mentor(s): arlos Celemin (TU Delft), Jens Kober (TU Delft) - Sjouke de Zwart
"Impact-Aware Learning from Demonstration"
Mentor(s): Alessandro Saccon (TU/e), Jens Kober (TU Delft) - Julian Freiherr von der Goltz
"Classification of Damages on Aircraft Inspection Images Using Convolutional Neural Networks"
Mentor(s): Wei Pan (TU Delft), Jochem Verboom (Mainblades Inspections), Jens Kober (TU Delft) - Thijs Greevink
"Prioritized Experience Replay based on the Wasserstein Metric in Deep Reinforcement Learning"
Mentor(s): Tim de Bruin (TU Delft), Jens Kober (TU Delft) - Jan Scholten
"Deep Reinforcement Learning with Feedback-based Exploration"
Mentor(s): Carlos Celemin (TU Delft), Jens Kober (TU Delft) - Daan Wout
"Policy Learning with Human Teachers - Using directive feedback in a Gaussian framework"
Mentor(s): Carlos Celemin (TU Delft), Jens Kober (TU Delft) - Mart Beeftink
"Learning Kinematic Models using a Single Tele-Demonstration"
Mentor(s): Ewout Arkenbout (HIT), Jens Kober (TU Delft) - Sathish Krishnamoorthi
"Model-Based Compensation for Serial Manipulators through Semi-Parametric Gaussian Process Regression"
Mentor(s): Jens Kober (TU Delft) - Martijn Kist
"Development of sway reduction control for the Jacket Lift System of Pioneering Spirit"
Mentor(s): Niels Mallon (Allseas), Marijn Dijk (Allseas), Jens Kober (TU Delft) - Konstantinos Kokkalis
"LMI-based Stability Analysis for Learning Control"
Mentor(s): Sebastian Trimpe (MPI-IS), Jens Kober (TU Delft) - Bart Keulen
"Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning"
Mentor(s): Jerry Pratt (IHMC), Jens Kober (TU Delft) - Wuyang Duan
"Learning State Representations for Robotic Control - Information Disentangling and Multi-modal Learning"
Mentor(s): Jens Kober (TU Delft) - Jovana Radojevic
"Cooperative Visual Object Learning"
Mentor(s): Lydia Fischer (HRIeu), Heiko Wersing (HRIeu), Stephan Hasler (HRIeu), Jens Kober (TU Delft) - Nikol Guljelmovic
"Task Parameter Inference in Human-Robot Interaction"
Mentor(s): Martijn Zeestraten (IIT), Sylvain Calinon (IIT), Jens Kober (TU Delft), Pieter Jonker (TU Delft) - Siddharth Ravi
"Reinforcement Learning Across Timescales"
Mentor(s): Tim de Bruin (TU Delft), Jens Kober (TU Delft) - Divyam Rastogi
"Deep Reinforcement Learning for Bipedal Robots"
Mentor(s): Ivan Koryakovskiy (TU Delft), Jens Kober (TU Delft) - Ronald Olsthoorn
"Segmented Active Reward Learning"
Mentor(s): Jens Kober (TU Delft), Robert Babuška (TU Delft) - Yannick Kathmann
"Multi-Sensor Human Detection on an Intelligent Security Robot - A study on combining vision and laser scan data"
Mentor(s): Jochem Verboom (RSS), Eelko van Breda (RSS), Floris Gaisser (TU Delft), Jens Kober (TU Delft), Pieter Jonker (TU Delft) - Niels Marck
"Semi-parametric Identification of Manipulator Dynamics in a Time-varying Environment"
Mentor(s): Jens Kober (TU Delft) - Derk Mus
"Generating Robotic Manipulation Trajectories with Neural Networks"
Mentor(s): Zoltán Szlávik (IBM), Robert-Jan Sips (IBM), Jens Kober (TU Delft) - Xiang Zhang
"Learning Control Policies for Tooth Extraction - An Application of Programming by Demonstration"
Mentor(s): Jens Kober (TU Delft), Tom van Riet (AMC) - Victor van Spaandonk
"Learning Variable Impedance Control: A Model-Based Approach Using Gaussian Processes"
Mentor(s): Jens Kober (TU Delft) - Renée van der Wijden
"Preference-driven Demonstration Ranking for Inverse Reinforcement Learning"
Mentor(s): Corrado Grappiolo (Alten), Chris Kalis (Alten), Jens Kober (TU Delft), Ivan Koryakovskiy (TU Delft), Martijn Wisse (TU Delft) - Jelle Munk
"Deep Reinforcement Learning - Pretraining actor-critic networks using state representation learning"
Mentor(s): Jens Kober (TU Delft), Robert Babuška (TU Delft) - Denise Feirstein
"Reinforcement Learning of Potential Fields to achieve Limit-Cycle Walking"
Mentor(s): Heike Vallery (TU Delft), Ivan Koryakovskiy (TU Delft), Jens Kober (TU Delft) - Jihong Zhu
"Real-time Collision Avoidance for Robot Manipulators by Braking Trajectory Prediction"
Mentor(s): Nico Mansfeld (DLR), Jens Kober (TU Delft) - Tamis van der Laan
"Consolidated Deep Actor Critic Networks"
Mentor(s): Marco Loog (TU Delft), Jens Kober (TU Delft)