Topics of student projects /  Témata studentských prací

[EN]: Below, you will find a list of all currently available topics for various university projects. For english speaking students: Some of the topics' details are written in Czech. However, do not be afraid to contact the supervisors for more information about the projects.

[CZ]: Některá témata jsou psána v angličtině a jiná v češtině. Nebojte, všechna témata je možné vypsat v obou jazycích. V případě zájmu o téma kontaktujte konkrétního vedoucího. 

Student must subscribte to a topic via the system: https://hub.fel.cvut.cz/topics/semestral_projects/all_topics_semestral

The topics managed by MRS group can be listed under department "13167". 

 

Agile Agile drone flight control refers to the ability of unmanned aerial vehicles (UAVs) to perform fast, precise, and adaptive maneuvers in dynamic environments. It involves advanced control algorithms that enable rapid response to external disturbances, such as wind or moving obstacles, while maintaining stability and accuracy. This field combines elements of control theory, robotics, and real-time computing to achieve high-performance autonomous flight.
High level planning and motion planning Robotic motion planning is the process of determining a sequence of movements that a robot must take to reach a goal without colliding with obstacles. It involves algorithms that compute feasible and efficient paths in the robot’s environment, considering its physical constraints. High-level motion planning extends this concept by incorporating task-level objectives, such as deciding which actions to perform and in what order to achieve a broader goal. It often involves reasoning about goals, environment changes, and multiple possible strategies. Together, these approaches enable robots to operate autonomously in complex, real-world scenarios.
Localization Robot localization is the process of determining a robot’s position and orientation within a known environment, which is essential for autonomous navigation and task execution. It typically uses sensor data such as GPS, cameras, lidar, or inertial measurements to estimate the robot’s location relative to a map or coordinate system. For drones (UAS), localization presents additional challenges due to their operation in 3D space and the need for precise positioning during fast, agile flight. GPS signals can be unreliable or unavailable in indoor or urban environments, making alternative methods like visual odometry, SLAM (Simultaneous Localization and Mapping), or sensor fusion critical. Drones also face issues with sensor drift, latency, and limited onboard computational resources. Maintaining accurate localization in real-time, especially during high-speed maneuvers or in dynamic environments, remains a complex and active area of research.
Drone Swarming Drone swarming refers to the coordinated control of multiple drones working together to achieve a shared objective, often inspired by the collective behavior of animals like birds or insects. In a swarm, each drone operates autonomously while communicating with others to maintain formation, avoid collisions, and adapt to changes in the environment. Swarming enables scalable and robust operations, such as area surveillance, search and rescue, or environmental monitoring. The underlying algorithms often involve decentralized decision-making, local sensing, and simple interaction rules that lead to complex, intelligent group behavior. This approach enhances efficiency, fault tolerance, and the ability to cover large or dynamic areas.
Applications Drones are widely used in agriculture for crop monitoring, spraying, and assessing field health using aerial imagery. In disaster response, they assist in search and rescue missions by quickly reaching areas that are unsafe or inaccessible to humans. Additionally, drones are employed in infrastructure inspection, such as monitoring bridges, power lines, and pipelines, to detect damage without risking human safety. This topic gathers tasks related to our industrial applications of flying robots.



Agile



Plánování letu bezpilotní helikoptéry pomocí posilovaného učení

Cílem projektu je návrh metod posilovaného učení pro plánování pohybu bezpilotního vzdušného prostředku.
Klasické metody plánování a řízení dronu nedokáží využít plný potenciál dronů při letu vysokou rychlostí skrze prostředí s překážkami.
V rámci projektu bude úkolem využít metody strojového učení, jako například posilované učení, pro zlepšení plánování a řízení dronu v neznámém prostředí.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Trajectory generation for agile flight in cluttered environment with the MRS system

The task In this project is to integrate the trajectory planning method [3] of agile quadrotor flight into Robot Operating System (ROS) so that this planner can be used during flight using the MRS system [2]. Furthermore, the planner will be extended by the possibility of using a grid map which can be created during the flight from on-board sensors. The planner will be tested on a realistic scenario of agile simulated quadrotor flight combined with appropriate predictive control method.
[1] Romero, Angel, Robert Penicka, and Davide Scaramuzza. "Time-optimal online replanning for agile quadrotor flight." IEEE Robotics and Automation Letters 7.3 (2022): 7730-7737.
[2] Baca, T., Petrlik, M., Vrba, M. et al. The MRS UAV System: Pushing the Frontiers of Reproducible Research, Real-world Deployment, and Education with Autonomous Unmanned Aerial Vehicles. J Intell Robot Syst 102, 26 (2021).
[3] Krystof Teissing, “Replanning of Collision-Free Trajectories for Unmanned Aerial Vehicle,” Master thesis, Czech Technical University in Prague, 2023.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Kinematic Traveling Salesman Problem (KTSP) with minimum time objective for unmanned aerial vehicle

The goal of this project is to design and implement an algorithm for solving the Kinematic Traveling Salesman Problem (KTSP) for multirotor UAVs similar to the Orienteering Problem in [1] . The solution will involve optimizing the sequence of waypoint fly-throughs to minimize time while taking into account the UAV’s simplified point-mass model [2]. The combinatorial part of the TSP will be solved with a heuristic such as Variable Neighborhood Search (VNS) method [3] or similar to the one in [1]. The results will be validated on existing TSP datasets and with UAV flight in simulations.
Literatura
[1] F. Meyer and K. Glock, "Kinematic Orienteering Problem With Time-Optimal Trajectories for Multirotor UAVs," in IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 11402-11409, 2022. [2] Teissing, K., Novosad, M., Penicka, R., & Saska, M. (2024). Real-time Planning of Minimum-time Trajectories for Agile UAV Flight. IEEE Robotics and Automation Letters, 1–8. https://doi.org/10.1109/lra.2024.3471388 [3] Robert Pěnička, Jan Faigl, Martin Saska and Petr Váňa. Data collection planning with non-zero sensing distance for a budget and curvature constrained unmanned aerial vehicle. Autonomous Robots 43(8):1937–1956, December 2019.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Energy-Efficient Model Predictive Contouring Control for UAVs

The goal of this project is to develop and analyze a Model Predictive Contouring Control (MPCC) strategy for UAVs that prioritize energy efficiency. While traditional MPCC methods focus on minimizing contouring and lag errors along a reference path while maximizing speed, they do not explicitly consider energy consumption. In real-world applications, optimizing energy use is critical for extending flight time and improving UAV performance.
Literatura
A. Romero, S. Sun, P. Foehn and D. Scaramuzza, "Model Predictive Contouring Control for Time-Optimal Quadrotor Flight," in IEEE Transactions on Robotics, vol. 38, no. 6, pp. 3340-3356, Dec. 2022, doi: 10.1109/TRO.2022.3173711. D. Lam, C. Manzie and M. Good, "Model predictive contouring control," 49th IEEE Conference on Decision and Control (CDC), Atlanta, GA, USA, 2010, pp. 6137-6142, doi: 10.1109/CDC.2010.5717042.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Autonomous flight of multiple drones using Imitation Learning

This project will aim to design a simulation pipeline for learning to fly multiple Unmanned Aerial Vehicles (UAVs) using imitation learning. The student will first study existing approaches to imitation learning and then select a suitable multi-UAV planning algorithm to imitate. Finally, the learned policy shall be shown to fly at least two UAVs inside a simulated environment.
Literatura
[1]R. Penicka, Y. Song, E. Kaufmann and D. Scaramuzza, "Learning Minimum-Time Flight in Cluttered Environments," in IEEE Robotics and Automation Letters, vol. 7, no. 3, pp. 7209-7216, July 2022, doi: 10.1109/LRA.2022.3181755. [2]A. Gleave et al., “imitation: Clean Imitation Learning Implementations”, https://arxiv.org/abs/2211.11972 [2]Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction. MIT press. [3]Raffin, Antonin, et al. "Stable-baselines3: Reliable reinforcement learning implementations." Journal of Machine Learning Research (2021).

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Minimum-time trajectory planning for autonomous drones

The goal of this project is to design a trajectory planning algorithm that minimizes the flight time of a drone and is able to replan the trajectory during flight. The task is to implement one of the trajectory planning methods presented in [2,3] and compare its results with the existing method [1]. Furthermore, the student will propose a way to extend the implemented method to include planning that takes into account an environment with obstacles.
Literatura
[1] Foehn P, Romero A, Scaramuzza D. Time-optimal planning for quadrotor waypoint flight. Science Robotics. 2021 Jul 21;6(56):eabh1221. [2] Fork T, Borrelli F. Euclidean and non-Euclidean Trajectory Optimization Approaches for Quadrotor Racing. arXiv preprint arXiv:2309.07262. 2023 Sep 13. [3] Qin C, Michet MS, Chen J, Liu HH. Time-Optimal Gate-Traversing Planner for Autonomous Drone Racing. arXiv preprint arXiv:2309.06837. 2023 Sep 13.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Learning-based multi-goal path planning

The goal of this project is to design a learning-based method for planning paths over multiple targets solving, at least partially, the orienteering problem [1]. The project will focus on designing either an algorithm with or without a teacher, so that the learned neural network can create an effective plan for visiting specified goals [1], or is able to make the heuristic search for solutions more efficient [4]. As part of the project, the student will: 1) study existing methods for path planning over multiple targets that use machine learning, 2) propose a suitable learning-based method that will make the search for a solution to the kinematic variant of the orientation problem more effective [2], 3) compares the proposed method with existing approaches.
Literatura
[1] Aldy Gunawan, Hoong Chuin Lau, Pieter Vansteenwegen, Orienteering Problem: A survey of recent variants, solution approaches and applications, European Journal of Operational Research, Volume 255, Issue 2, 2016. [2] F. Meyer and K. Glock, "Kinematic Orienteering Problem With Time-Optimal Trajectories for Multirotor UAVs," in IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 11402-11409, Oct. 2022. [3] Li, Bingjie, et al. "An overview and experimental study of learning-based optimization algorithms for the vehicle routing problem." IEEE/CAA Journal of Automatica Sinica 9.7 (2022): 1115-1138. [4] Xin L, Song W, Cao Z, Zhang J. NeuroLKH: Combining deep learning model with Lin-Kernighan-Helsgaun heuristic for solving the traveling salesman problem. Advances in Neural Information Processing Systems. 2021 Dec 6;34:7472-83. [5] Kool, Wouter, Herke Van Hoof, and Max Welling. "Attention, learn to solve routing problems!." arXiv preprint arXiv:1803.08475 (2018).

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Nonlinear model predictive control for drone flight in cluttered environment

The aim of the project is to design model predictive control methods for drones in environments with obstacles. Classical control methods usually do not consider obstacles, and therefore, when following a collision-free reference trajectory, they may encounter obstacles in an effort to minimize steering deviation. This project will investigate predictive control methods that consider obstacles during flight.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Using learning-based methods for heuristics in routing problems

The goal of this project is to design a learning-based method to be used as a heuristic [1] in finding solutions to routing problems [2]. The project will focus on the design of algorithms with the teacher so that the learned neural network can estimate the cost of the trip and the states of the robot in the visited cities for a variant of routing problems with a kinematic robot. The proposed algorithm will be compared on existing datasets against the used teacher [3].
Literatura
[1] Li, Bingjie, et al. "An overview and experimental study of learning-based optimization algorithms for the vehicle routing problem." IEEE/CAA Journal of Automatica Sinica 9.7 (2022): 1115-1138. [2] Xin, Liang, et al. "NeuroLKH: Combining deep learning model with Lin-Kernighan-Helsgaun heuristic for solving the traveling salesman problem." Advances in Neural Information Processing Systems 34 (2021): 7472-7483. [3] Meyer, Fabian, and Katharina Glock. "Kinematic orienteering problem with time-optimal trajectories for multirotor uavs." IEEE Robotics and Automation Letters 7.4 (2022): 11402-11409. [4] Kool, Wouter, Herke Van Hoof, and Max Welling. "Attention, learn to solve routing problems!." arXiv preprint arXiv:1803.08475 (2018).

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Detection and localization of powerline insulators using lidar and camera data synthesis onboard unmanned aerial vehicle

During the project, the student will design a method for synthesizing lidar and camera data [1,2] to detect and locate insulators on a high voltage transmission line pole for visual inspection of the insulators [3]. The output data from the camera and lidar mounted on the drone are used to detect the insulators. Based on the localization, the algorithm is able to determine a suitable position to obtain an inspection image of the detected object. The design of the algorithm should be fast enough for later integration in a simulated or real environment using only the available computational capacity on the UAV.
Literatura
[1] Liu, H., Wu, C. & Wang, H. Real time object detection using LiDAR and camera fusion for autonomous driving. Sci Rep 13, 8056 (2023). https://doi.org/10.1038/s41598-023-35170-z [2] Mousa-Pasandi, M., Liu, T., Massoud, Y. et al.RGB-LiDAR fusion for accurate 2D and 3D object detection. Machine Vision and Applications 34, 86 (2023). [3] D. Sadykova, D. Pernebayeva, M. Bagheri and A. James, "IN-YOLO: Real-Time Detection of Outdoor High Voltage Insulators Using UAV Imaging," in IEEE Transactions on Power Delivery, vol. 35, no. 3, pp. 1599-1601, June 2020.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Gate Detection in drone racing using convolutional neural networks

The goal of this project is to design and implement an algorithm for gate detection using convolutional neural networks (CNN) to navigate autonomous drones. The system will be capable of detecting gate corners and estimating their position based on known geometrical properties of the gates and their location. Part of the work is also the creation of sufficient data sets for learning detections. The resulting algorithm should be robust enough to handle partial visibility of gates or overlapping gates. Functionality will be tested in a simulated environment and subsequently in a real-world setting.
Literatura
[1] Foehn, Philipp, et al. "Alphapilot: Autonomous drone racing." Autonomous Robots 46.1 (2022): 307-320. [2] Kaufmann, Elia, et al. "Champion-level drone racing using deep reinforcement learning." Nature 620.7976 (2023): 982-987. [3] D. Hanover et al., "Autonomous Drone Racing: A Survey," in IEEE Transactions on Robotics, vol. 40, pp. 3044-3067, 2024.

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Reinforcement learning for trajectory planning of drones in a cluttered environment

The project aims to design reinforcement learning methods for unmanned aerial vehicle trajectory planning. Classical drone planning and control methods fail to exploit the full potential of drones in high-speed flight through cluttered environments. The project will focus on the use of machine learning methods, such as reinforcement learning, to improve drone planning and control in unknown environments.
Literatura
[1]R. Penicka, Y. Song, E. Kaufmann and D. Scaramuzza, "Learning Minimum-Time Flight in Cluttered Environments," in IEEE Robotics and Automation Letters, vol. 7, no. 3, pp. 7209-7216, July 2022, doi: 10.1109/LRA.2022.3181755. [2]Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction. MIT press. [3]Raffin, Antonin, et al. "Stable-baselines3: Reliable reinforcement learning implementations." Journal of Machine Learning Research (2021).

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Nonlinear Model Predictive Control for multi-robot aerial vehicles with Boids flocking model

The student will design a nonlinear model predictive control (NMPC) such that a multi-robot system composed of multiple unmanned aerial vehicles (drones) can fly in a swarm using the Boids model [1], which will be integrated within the NMPC. Robots should not collide during flight, but at the same time, robots should not be too far apart and should be able to fly in a swarm towards a common goal. The control functionality will be verified in the simulation.
Literatura
[1] Baca, T., Petrlik, M., Vrba, M., Spurny, V., Penicka, R., Hert, D., & Saska, M. (2021). The MRS UAV system: Pushing the frontiers of reproducible research, real-world deployment, and education with autonomous unmanned aerial vehicles. Journal of Intelligent & Robotic Systems, 102(1), 26. [2] Reynolds, C. W. (1987, August). Flocks, herds and schools: A distributed behavioral model. In Proceedings of the 14th annual conference on Computer graphics and interactive techniques (pp. 25-34).

Advisor: Robert Penicka
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




High level planning and motion planning



Disassembly path planning

The task of disassembly planning is to disassemble an object consisting of several parts.
On the highest level, disassembly sequencing finds subsequent disassembly actions that can separate individual parts of an assembly.
This problem is usually formulated as a discrete search and optimization problem.
Disassembly path planning deals with computing collision-free paths (motions) for the individual parts.
The task of this project is to:
a) investigate state-of-the-art in the assembly or disassembly planning
b) implement (or use) suitable solver for simplified disassembly planning considering that only one piece can move at time
c) implemented suitable path planner to achive the movements of the objects.

Advisor: Vojtech Vonasek
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Space-filling trees in sampling-based motion planning

Motion planning of robots/other objects leads to a search in high-dimensional configuration spaces.
This is typically solved by sampling-based planning, where random samples are drawn (usually uniformly) in the space, classified as free/non-free and the free ones are stored into a roadmap.
Path in the roadmap then describes motion of the robot/object.
The way of sampling influences the performance of the planning (speed, efficiency, quality..).
The goal of this project is to investigate usage of space-filling trees for sampling-based planning.

Advisor: Vojtech Vonasek
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.



Localization



Long-term self-localization of UAVs in mapped environments

Develop an algorithm for precise 3D self-localization of an UAV in an a priori known map of the environment. The UAV will be equipped with a GPS receiver and a 3D LiDAR sensor, so an ICP-inspired approach is a suitable solution. The map should be updated online by the algorithm to account for changes in the environment (such as a parked car, which is missing in the a priori map, or a tree, which was cut down). Evaluate the precision, robustness and general performance of your algorithm in a real-world experiment with flying UAVs. This topic is motivated by the task of periodic autonomous inspection of infrastructure by UAVs

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Gimbal detection and tracking for a small autonomous UAV

Use a detection algorithm to find an object of interest in an image and control a camera gimbal to track the object. The tracking has to be robust to rapid movements of the UAV, carrying the gimbal. The gimbal also has zoom capabilities, which the control algorithm should take into account to provide an optimal view of the target. Evaluate the precision, robustness and general performance of your solution in a real-world experiment with flying UAVs. Motivation of this topic is autonomous monitoring of workers in high-risk environments for safety purposes and autonomous tracking of high-speed targets for the purpose of physical interaction or collision avoidance

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Drone detection and relative localization from a thermal camera

Develop a detection and relative localization algorithm for detection of drones using a thermal camera, placed onboard a flying UAV. The detection task may be tackled using a convolutional neural network. A combination of thermal and RGB cameras for the detection input is also possible. The resulting algorithm should provide good precision and low latency to be used for the task of autonomous drone interception (see http://mrs.felk.cvut.cz/projects/eagle-one). Evaluate precision, detection range and general performance of the algorithm in a real-world experiment with several flying UAVs.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Coordination of a Heterogeneous Team of Human-Aerial Co-Worker (ACW)

The main objective of this task will be to simultaneously provide continuous assistance and monitoring of a person working in a hazardous environment by a group of cooperating aerial co-workers. A crucial information required for safe and efficient group coordination is a reliable knowledge of the states of ACWs and also of the relative positions with respect to the human workers. To provide a knowledge of the full state of the group, a distributed fusion mechanism will be designed using outputs of an onboard relative visual localization system between the ACWs and relative to the humans.
Thus, an advantage of localization sensors distributed among the entire group will be exploited to increase precision and reliability of the overall process. In addition, requirements of the pose estimator will be embodied into the multi-objective group coordination to increase the reliability and precision of the system. This topic is motivated by a large European project on autonomous powerline inspection by a team of UAVs

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Drone detection using convolutional neural networks

Use a convolutional neural network for detection of drones using an RGB camera, placed onboard a flying UAV. It is possible to utilize automatic dataset annotation using the UVDAR system for training of the CNN (seehttp://mrs.felk.cvut.cz/projects/midgard). Implement the whole solution to run online, onboard the UAV. The resulting algorithm should provide good precision and low latency to be used for the task of autonomous drone interception (see http://mrs.felk.cvut.cz/projects/eagle-one). Evaluate precision, detection range and general performance of the algorithm in a real-world experiment with several flying UAVs.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Drone detection using neural networks and LiDAR

Use a neural network for detection of drones using a 3D LiDAR sensor, placed onboard a flying UAV. It is possible to utilize automatic dataset annotation using the UVDAR system for training of the neural network (seehttp://mrs.felk.cvut.cz/projects/midgard). Implement the whole solution to run online, onboard the UAV. The resulting algorithm should provide good precision and low latency to be used for the task of autonomous drone interception (see http://mrs.felk.cvut.cz/projects/eagle-one). Evaluate precision, detection range and general performance of the algorithm in a real-world experiment with several flying UAVs.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Intelligent automatic camera exposure control for robot sensing

A correct setting of exposure duration and gain of a camera strongly influences the quality of information in the resulting captured image. Even though, this problem is often delegated to naïve or proprietary algorithms integrated in the respective camera, that are not sufficiently robust for general robot perception. Research and compare state-of-the-art intelligent methods of automatic exposure and gain control for cameras. Select and implement a suitable solution for deployment onboard unmanned aerial vehicles. Experimentally validate the implemented solution.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Multi-target tracking for unmanned aerial vehicles

Research and compare algorithms for multi-target tracking. Select a suitable algorithm for deployment on an onboard computer of an unmanned aerial vehicle in the task of autonomous tracking of an unknown number of targets. Implement and experimentally evaluate the selected algorithm.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


3D reconstruction from a monocular camera image using CNN features

Algorithms for 3D reconstruction of environment using a monocular camera (SLAM, SfM etc.) have many uses in mobile robotics. When deployed on UAVs, these can be used eg. for inspection, mapping or obstacle avoidance. The basis of most of these algorithms is a detector of visual features that are then used to estimate the camera's movement, position of obstacles in the environment etc. Lately, several feature extraction algorithms based on deep learning have been published that promise better robustness and accuracy than conventional approaches. Implement an algorithm for detection of visual features in a camera image using neural networks so that it can run on an onboard computer of a UAV with minimal latency. Test the feature detector and compare it with the ORB and SIFT conventional detectors for the application of 3D environment reconstruction.

Advisor: Matous Vrba
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Drone Swarming



Fast formation flight through cluttered environment using formation shape adaptation

The goal of this project is to design and implement algorithm for safe navigation of formation of UAVs through environment with obstacles by adapting the shape of the formation. The project consists of two tasks. First part will focus on modification of an existing algorithm [1] for formation shape adaptation for the use in curvilinear coordinate systems. Second part of the project will focus on development of algorithm for planning of a sequence of formation shapes that will lead to a safe navigation of the formation along given path in 3D environment. The developed algorithm is expected to be deployed on a fleet of UAVs in a real-world experiment.

Reference [1] V. Kratky, R. Penicka, J. Horyna et al., "CAT-ORA: Collision-Aware Time-Optimal Formation Reshaping for Efficient Robot Coordination in 3-D Environments," in IEEE Transactions on Robotics, vol. 41, pp. 2950-2969, 2025.

Advisor: Vit Kratky
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Formation shape adaptation in environments with obstacles

The goal of this project is to design and implement algorithm for formation shape adaptation in environments with obstacles with focus on time efficiency of the formation reshaping process. The designed algorithm is expected to be composed of solution to both the robot-to-goal assignment problem and trajectory generation problem in environments with obstacles. The developed algorithm will be tested in real-world experiments with unmanned aerial vehicles.

Advisor: Vit Kratky
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.



Applications



Autonomní dron Ryze Tello / Autonomous dron Ryze Tello

Firma DJI společně s firmou Intel vyvinula unikátní dron Ryze Tello, který je připraven pro ovládání z počítače přes wifi připojení.
Cílem práce je vytvořit systém, který na základě informací z kamery a výškoměrů bude schopen orientovat se v uzavřeném prostředí a dovede se naučit cesty.

Advisor: Petr Stepan
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Skupina autonomních dronů Ryze Tello / Group of autonomous drones Ryze Tello

Seznamte se s ovládáním dronu Ryze Tello a robotickým operačním systémem ROS.
Seznamte se s algoritmy pro rojové chování dronů.
Připravte SW pro ovládání více dronů Ryze Tello.
Dále s pomocí výsledků navazujících prací implementujte skupinové chování, synchronizované provedení předdefinovaných plánů.

Advisor: Petr Stepan
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Vytvoření modelu stožáru elektrického vedení dronem s termální kamerou / Modelling of power line pylon using drone with thermal camera

Drony jsou v poslední době hojně využívány k inspekci stožárů elektrického vedení. Cílem této práce je vytvořit model stožáru elektrického vedení z drony, která se pohybuje okolo stožáru. Prá se může zaměřit na vytváření modelu z obrazu stereo barevných kamer, z obrazu stereo termálních kamer, nebo s monokulárního systému při známé pozici drony. Téma je vhodné i pro více studentů, neboť je možné využít různé senzory pro tvorbu modelů a jejich výsledky pak vzájemně porovnat.

Advisor: Petr Stepan
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Navigace a lokalizace dronu s termokamerou / Drone localization and navigation with thermal camera

Termokamery poskytují nový pohled na okolí, které nás obklopuje. I pro roboty je využití termokamer výhodné, protože umožňuje detekovat objekty, které v obyčejné kameře jsou špatně viditelné. Cílem této práce je vytvořit algoritmy pro detekci objektů v termokameře, nebo v páru stereo termokamer. Pomocí detekce těchto známých předmětů (jako je stožár elektrického vedení, samotné elektrické vedení) navigovat dron při průzkumu stožáru, případně při přelétání mezi stožáry.

Advisor: Petr Stepan
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Assistive technology for scanning and documentation of historical monuments by an autonomous helicopter

Cílem práce je návrh a vývoj asistivní technologie pro operátora bezpilotní helikoptéry řešící úlohu senzorického snímání a fotografování obtížně přístupných míst v interiérech a exteriérech rozlehlých historických budov (kostelů, hradů, zřícenin).
Výstupem projektu bude systém umožňující operátorovi helikoptéry vizualizovat informaci o bezprostředním okolí helikoptéry (vzdálenost k překážkám) a navrhovat korekční úhybné manévry, což umožní získat informaci z míst, které není možné dokumentovat běžnou technologií.
Cílem práce je odvážně se pouštět (s helikoptérou) tam, kam se dosud (s moderními senzory) nikdo nedostal.
V případě úspěšně realizovaného systému bude volitelně na projekt navazovat honorovaná práce s historiky v sakrálních objektech Středočeského a Olomouckého kraje v rámci projektu: http://mrs.felk.cvut.cz/projects/cesnet.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Autonomous cooperative object gathering by a group of cooperating helicopters

Cílem práce bude navrhnout, implementovat a experimentálně ověřit systém pro řízení formace helikoptér s vizuální zpětnou vazbou, který umožní autonomně uchopit relativně lokalizované předměty a stavit z nich zeď.
Student bude během práce řešit identifikaci modelu helikoptéry, řešit plánování pohybu a navrhovat vhodné zpětnovazební řízení a ovládání palubního mechanismu pro uchopení jednotlivých předmětů.
Práce bude směřovat k účasti na mezinárodní soutěži v Abu Dhabi, kde budeme opět soutěžit s nejlepšímy týmy světa http://mrs.felk.cvut.cz/projects/mbzirc.
Součástí úspěšné realizace práce je možnost stáže na jednom z nejlepších robotických pracovišť světa GRASP lab university v Pennsylvanii.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Formations of relatively stabilized helicopters in transmission source localization tasks

Cílem práce je integrovat principy měření vzdálenosti vysílače a přijímače neseného helikoptérou z intenzity přijímaného signálu do systému řízení formace relativně stabilizovaných bezpilotních helikoptér vyvíjeného skupinou Multi-robotických systémů Katedry kybernetiky a využít je v úloze kooperativní lokalizace čipu.
Poloha čipů v prostoru bude lokalizována triangulací měření z minimálně trojice helikoptér a Kalman filtrací získaných dat obsahujících šum.
Předpokládá se kombinace práce s HW (instalace přijímače a jeho integrace se systémem pro řízení helikoptéry) a SW (implementace navrženého systému pro filtraci dat, odhad polohy předmětů a plánování pohybu helikoptér s cílem tyto polohy zpřesnit).
Při řešení projektu se předpokládá úzká spolupráce s průmyslovým partnerem skupiny a v případě úspěšné realizace možnost návazných prací pro tuto firmu.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Mechanism for objects manipulation by an unmanned helicopter

Cílem práce bude navrhnout a vyvinout inteligentní zařízení nesené helikoptérou a propojené se systémem jejího řízení pro úlohu autonomního uchopování a přemísťování předmětů, kterou bude skupina Multi-robotických systémů Katedry kybernetiky řešit v rámci soutěže MBZIRC http://mbzirc.com/.
Student bude mít za úkol navrhnout a realizovat konstrukční design zařízení a elektroniku pro jeho autonomní ovládání palubními systémy helikoptéry a pro detekci správného uchopení předmětu.
Volitelnou součástí práce bude senzorická fúze dat ze zařízení a dalších senzorů helikoptéry pro odhad přesnosti uchopení předmětu.
V případě úspěšně realizovaného systému bude možné zúčastnit se přípravného kempu v USA a vlastní robotické soutěže ve Spojených arabských emirátech.
Pojďte s námi změřit síly v pouštním království s nejlepšími universitami světa. http://mrs.felk.cvut.cz/projects/mbzirc

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Multi-robot surveillance by a group of unmanned helicopters and cooperative autonomous aircrafts

Cílem práce bude vyvinout systém pro plánování pohybu a koordinaci skupiny helikoptér a bezpilotních letounů tak, aby se vhodně zkombinovaly přednosti obou platforem.
Helikoptéry dokáží manévrovat v malých rychlostech blízko překážek, zatímco letouny profitují z vyšší maximální rychlosti, většího doletu a delší operační doby.
Navržený a implementovaný systém bude verifikován v robotickém simulátoru a v případě úspěšné realizace systému otestován s reálnými helikoptérami skupiny Multi-robotických systémů Katedry kybernetiky.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




System of autonomous localization and fire extuinguishment by a drone in tall buildings

Cílem práce bude navrhnout, implementovat a experimentálně ověřit systém pro řízení dronu, který umožní autonomně lokalizovat ohniska požáru.
Student si bude moci vybrat řešení SW části úlohy, řešení HW části úlohy a nebo kombinaci obojího.
SW část bude obsahovat plánování pohybu, řízení a stabilizaci bezpilotní helikoptéry.
HW část se bude sestávat z návrhu mechanismu pro vlastní hašení a jeho integraci na palubě dronu.
Práce bude směřovat k účasti na mezinárodní soutěži v Abu Dhabi, kde budeme opět soutěžit s nejlepšímy týmy světa http://mrs.felk.cvut.cz/projects/mbzirc.
Součástí úspěšné realizace práce je možnost stáže na jednom z nejlepších robotických pracovišť světa GRASP lab university v Pennsylvanii nebo University of New York.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Motion planning of a group of helicopters in autonomous construction task

Cílem práce je vyvinout systém pro plánování a koordinaci skupiny bezpilotních helikoptér v úloze autonomního kooperativního sběru statických s cílem stavět z nich konstrukci.
Student na základě poskytnuté neurčité informace o poloze objektů navrhne algoritmus založený na principech robotického pokrytí a TSP pro plánování bezkolizních trajektorií helikoptér s cílem získat snímky vytipovaných oblastí ve vyšším rozlišení a následně naplánovat trajektorie do míst s potvrzeným výskytem objektů (vlastní autonomní sběr objektů není povinnou součástí práce).
Práce bude směřovat k účasti na mezinárodní soutěži v Abu Dhabi, kde budeme opět soutěžit s nejlepšímy týmy světa http://mrs.felk.cvut.cz/projects/mbzirc.
Součástí úspěšné realizace práce je možnost stáže na jednom z nejlepších robotických pracovišť světa GRASP lab university v Pennsylvanii.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




Bio-inspired autonomous robotic swarms

Seznamte se s algoritmy pro řízení a navigaci autonomních formací inspirovaných pohybem hejn ptáků, ryb či hmyzu.
Navrhněte a implementujte vhodný algoritmus pro ovládání a stabilizaci roje bezpilotních helikoptér.
Roj helikoptér by se měl být schopen pohybovat v dynamickém prostředí s překážkami a tvar roje by se měl automaticky přizpůsobovat řešeným úlohám mobilní robotiky.
Metodu ověřte a analyzujte pomocí simulací a dílčím reálným experimentem s bezpilotními helikoptérami skupiny Multi-robotických systémů. Předpoklady: základní znalost programování v C nebo Pythonu.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.


Moving object tracking by a group of relatively localized unmanned helicopters

Cílem práce je vyvinout systém pro automatické sledování pohybujícího se objektu skupinou helikoptér, ve kterém je informace o pozici helikoptér ve formaci a přesnost detekce objektu jednotlivými helikoptérami využita k řízení pohybu formace a zvýšení robustnosti lokalizace.
Hlavní náplní bude implementace a experimentální ověření (v simulátoru i s reálnými helikoptérami) metody pro řízení formace a integrace palubních senzorů a systému helikoptér.
Práce navazuje na spolupráci skupiny Multi-robotických systémů Katedry kybernetiky a pracoviště Queen Mary University of London a bude vyžadovat koordinaci a komunikaci s kolegy z Londýna.
V případě úspěšné realizace systému lze dohodnout stáž na partnerském pracovišti.

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.




System for intelligent photography and filming by a group of helicopters

Cílem práce je integrovat principy a metody využívané profesionálními fotografy a filmaři do systému pro polo-autonomní dokumentaci obtížně přístupných míst v interiérech budov skupinou vzájemně spolupracujících helikoptér.
Výstupem práce bude rozšíření systému pro stabilizaci formace helikoptér vyvíjeného v rámci skupiny Multi-robotických systémů na Katedře kybernetiky, který umožní adaptivně nastavit relativní vzdálenost a úhel mezi kamerou nesenou jednou z helikoptér a reflektory na sousedních helikoptérách.
Téma je vhodné pro fotografické nadšence, kteří by chtěli svůj koníček využít ve studiu robotických systémů a jejich aplikací.
Součástí tohoto tématu bude možnost výcviku řízení bezpilotních helikoptér. http://mrs.felk.cvut.cz/projects/cesnet

Advisor: Martin Saska
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.