As the autonomy and AI technologies are evolving, the challenges in developing complete autonomous missions remain, especially in extreme environments with perceptual degradation, difficult terrain, high-risk operations and so on. The focus of this Master Thesis is towards identifying the key modules used for autonomous exploration and map building with multiple UAVs, and developing important modules that are currently missing or under-perform in extreme environment scenarios. One of them is the ability to merge multiple robots' 3D point cloud maps, in order to have a robust and efficient path planning algorithm. Another one is the global pose estimation algorithms in featureless and GPS-denied environments.
The initial work in this master thesis will be towards solving the fundamental problems with ERRT in speeding up the multi-path planner, and by adding integration through ROS with mapping and frontier generation modules, such that ERRT can directly take online map updates from an occupancy mapper such as OctoMap or VoxBlox.
Once the framework is operational, initial testing in real-time simulations such as Gazebo can be performed as a stepping stone towards real experiments.
The thesis can be progressed by the integration of multi-agent exploration behavior into the framework and by adding novel techniques such as adaptive sampling.
If you are interested to make a Master thesis with us please contact indicated persons in the available projects below or send an email to firstname.lastname@example.org. We would like to discuss your ideas with you and build a customised thesis that will fit your ambition 100%.
Evert Häggman - Development, Control and Investigation of a Robotic Exoskeleton for Astronaut Back Support
Niklas Dahlquist - Generation of Behavior Trees for Dynamical Agents Based on Market Based Task Allocation
Clement Petit - Mapping and Path planning for aerial vehicles in SubT environments
Jonathan Olsson - Detect Anomalies in Shapes Using Machine Learning Techniques
Alkis Sigkounas - Multi Agent Exploration with Reinforcement Learning
Enrico Giacomini - Design, Modeling and Control of a Thrust Based UAV
Vignesh Kottayam Viswanathan - Cooperative Navigation of Spacecrafts in proximity of small bodies
Tommaso Gasparetto - Docking mechanism design and development, detection and guidance for space robots
Eric Brune - AI empowered feature based detection and tracking for visual servoing of satellites
Scott Fredriksson - Design, Development and Control of a Quadruped Robot
Jerker Bergström - UAV path planning and collision avoidance in unknown environments
Matteo Terreran (Aalto Uni)- Machine Learning and Computer Vision for Aerial manipulation
Maros Hladky (Aalto Uni) - Vision Based Attitude Control
Issouf Ouattara (Aalto Uni) - Semi-autonomous drone System for Mapping and Measuring of Agricultural Fields and Forest Stands
Olivier Struckmeier (Aalto Uni) - Generating Explanations of Robot Policies in Continuous State Spaces
Usama Tariq - Robotic Grasping of Large Objects for Collaborative Manipulation
Anna Costalonga - Modeling and Control of a novel PMA enabled robotic arm
Elias Small - Tilt Rotor: Development and Control
David Wuthier (EPFL, Switzerland) - Aerial Manipulation
Nicola Dal Lago (University of Padova) - Aerial Mapping
Christoffer Carholt - Single Rotor UAV
Adrian Lindqvist - Modeling and Experimental Evaluation of a Tilt Rotor Aircraft
Rickard Nyberg - Development and Control of Unammaned Ground Vehicles (UGVs), now at VOLVO cars
Emil Fresk - Modeling, Control and Experimentation of a Variable Pitch Quadrotor
Ivan Monzon Catalan - Development of a ROS enabled Quadrotor
Theodoros Giannakas - Model Predictive Control of a Pulp & Paper Refiner