sada

Aerial-Ground cooperative systems

In recent years the interest in autonomous aerial vehicles (AUV) has had a massive increase. The research in this area has been focused on hardware design, modelling, control algorithms and more recently in aggressive maneuvers, collaborative tasks and the movement of swarms of UAV. One of the most interesting open problems in this field is the collaboration between autonomous aerial and ground robots, the main idea is to have a system of heterogeneous robots which work together in order to complete a goal, specifically in cases where the advantages of each robot complement the disadvantages of the others and allows them to achieve the goal, faster, more accurately or with a greater degree of safety.

In the Control Methods and Robotics Lab we currently have a team of Quadrotors and Omnidirectional robots with the required sensors for autonomous tasks and simulation environments based on V-REP, ROS and Gazebo. We are developing the next generation of algorithms for autonomous cooperative robotics and we offer several topics in this area. Feel free to ask or propose topics related to this research.

Autonomous UAV obstacle avoidance.

In this project a UAV (unmanned aerial vehicle) must fly in indoor environments without colliding with walls, objects or persons in the room. The UAV has 2 cameras, Wifi communication and necessary sensors for the control of the platform. The student must propose and implement a system which takes the information from the cameras and internal sensors to avoid colliding with unknown obstacles while following a predefined path. This system must take in consideration the dynamics of the platform and define the proper controAlled motion for obstacle avoidance. The initial phase of the project will be developed in a simulation environment already available and later in the actual robot.

Goals

  • Research and implementation of Computer Vision algorithm for the detection of obstacles.
  • Research and definition of a motion controller for obstacle avoidance.

Requirenments

  • Strong programming skills (C, C++)

Quadcopter Localization and Mapping

The aim of this topic is the study, design and programming of algorithms for localization and mapping of a Quadcopter in indoor environments. A specific objective of this work is to generate a map of our laboratory and hallways and testing of already available algorithms in this map. This topic will be developed in a robotic simulator for posterior testing in a real quadcopter.

Goals:

  • Implementation of a self localization system using the internal camera.
  • Use the estimated position and the camera information to build a map of the environment.

Requirements:

  • General knowledge (or interest) in mapping and localization algorithms.
  • C++

Aerial ground cooperative vision

In this project the student must merge the camera information from the quadcopter with the camera information from a ground robot (i.e Robotino). This system will use the information from both sources in order to obtain a better estimate of the position of both aerial and ground robots related to the environment and possibly form a map with this information. Also control and navigation strategies can be implemented which improve the behavior of the system. This project can be implemented in a simulation.

Requirements:

  • Computer Vision knowledge.
  • C++

Aerial ground cooperative exploration

A quadcopter and a group of ground robots must explore and environment as fast as possible. In this scenario it is necessary to evaluate in which cases it is favorable to deploy the quadcopter given that it's battery autonomy is less than 20 minutes. A control scheme is needed which allows the quadcopter to return to the ground robots even after losing camera contact with them (but maintaining communication). This project will be implemented in a robotic simulator.

Requirements:

  • Computer Vision knowledge.
  • Good C++ programming skills.

Running projects

Quadcopter velocity control with wifi signal delays

Master Thesis. March 2016 – August 2016

Autonomous UAV landing on a moving platform

Master Thesis. March 2016 – August 2016

UAV as mobile fiducial markers

Bachelor Thesis. March 2016 – July 2016

Finished Projects

Design of a wireless camera system for an UAV.

Bachelor Thesis. October 2015 – March 2016

zurück zur Übersicht der am Fachgebiet angebotenen Arbeiten