Beitrags-Archiv für die Kategory 'DA_NES'

Master Thesis: Energy-Aware Multi-Robot Exploration

Dienstag, 19. April 2016 14:33

Background

In mobile robotics a very prominent task is to explore an unknown environment. This can be applied for search and rescue missions after a natural disaster or extraterrestrial missions such as the exploration of planets. In such hostile environments the autonomy of the robots plays a crucial role in the mission success, as humans usually cannot intervene to aid the robots. For long lasting missions the robots therefore have to manage the available energy and possibly recharge their battery, e.g. at a docking station. This requires an estimation of the battery‘s SOC and the remaining distance the robot is still able to travel. To speed up the mission and to make it more reliable, multiple robots can work together in a team. With multiple robots coordination becomes necessary for recharging and exploration to avoid congestions and redundant exploration. Inspired by economics, coordination by market-based approaches like auctioning promise great flexibility and good performance.

Task

Goal of this thesis is the energy-aware coordination of multiple mobile robots during the exploration of unknown environments. The coordination should work fully autonomous and decentralized for flexibility and scalability. The work will be done using the Robot Operating System (ROS) which provides the basic functionality and allows a focus on the high-level concepts.

The thesis consists of three major steps:

  • Implementation of the coordination algorithm in ROS
  • Verification through simulations and practical experiments on different robot platforms
  • Performance evaluation through simulations in ROS

Interested students should have basic knowledge of the Linux operating system and the C++ programming language.

Further Information

For further details contact Micha Rappaport.

Download PDF: aau_multirobot_thesis

Share

Thema: DA_Mobile Systems, DA_NES, Diplomarbeit | Kommentare (0) | Autor:

Internship or Master Thesis: Real-Time Wireless Communication Industrie 4.0

Freitag, 31. Oktober 2014 12:16

Your responsibilities:
You will be a part of a research team at Robert Bosch GmbH working on a networking platform for connected manufacturing in future factories (Industrie 4.0). Your main tasks will include:

  • Evaluation of real-time operation systems for embedded wireless devices (IEEE 802.15.4, Bluetooth, WLAN)
  • Setting-up a testbed for a wireless network with real-time services
  • Design and implementation of service and packet scheduler

Your competencies and qualifications:

  • Student of Electrical Engineering, Computer Science, or similar
  • Good programming skills in C/C++
  • Solid Linux user with first programming experience of embedded devices (Arduino, OpenWRT, TinyOS, embedded Linux, etc.)
  • Good knowledge of (wireless) networking protocols (MAC and routing)
  • Well-organized, responsible, and independent working style

Date and location:

  • Start: Dezember 2014
  • Length: 6 months
  • Job location: Renningen

Reference no.: DE00327359. Apply here. Please attach your examination regulations and when indicated work permit and legal alien resident. For more information contact Dr. Marchenko. Tel.: +49-711-811- 24518.

 

Share

Thema: DA_Mobile Systems, DA_NES, Diplomarbeit, Institut, NES, PR_Mobile Systems, PR_NES, Praktikum | Kommentare (0) | Autor:

Open master thesis in cooperation with SKIDATA

Donnerstag, 12. Juni 2014 13:09

Skidata

The Pervasive Computing Group and the Institute of Networked and Embedded Systems (NES) is looking for a master thesis student. The thesis is carried out in collaboration with SKIDATA and deals with:

– Novel access solutions based on computer vision
– Computer vision solutions for low-power embedded devices
– In-depth evaluation of embedded visual sensing solutions

Additional information about the thesis, the requirements as well as contact details are available here.

 

Share

Thema: DA_NES, DA_Pervasive Computing, Diplomarbeit, Forschung, Institut, NES, Studierendenprojekt | Kommentare (0) | Autor:

Joint Master Thesis with SenseFly (Lausanne)

Mittwoch, 28. Mai 2014 13:19

As part of our SINUS research project (http://uav.aau.at), NES is offering a joint master thesis with SenseFly (http://www.sensefly.com) – a leading drone company located in Lausanne, Switzerland. Interested students are requested to contact Prof. Bernhard Rinner.

Project Description

senseFly Ltd, a spinoff of the Swiss Federal Institute of Technology (EPFL) in Lausanne, develops and commercialises mini-drones for civilian applications in the domain of 2D and 3D mapping. The maps generated with these drones are used for such applications as construction, mining logistics, forestry, conservation, crop monitoring or humanitarian aid. As part of its extensive research and development program, senseFly is looking for motivated engineers and programmers to develop, design, test and even release to market new hardware and software modules for our mini-drones. As an intern or master’s student, you will be integrated directly with our dynamic team of engineers and researchers working on new products on the cutting edge of aerial robotics. Projects have a duration of 6 months and can be tailored based on the competencies of the applicant.

The proposed thesis project involves the development of smart flight planning algorithms for swarms of fixed-wing eBee drones (https://www.sensefly.com/drones/ebee.html). Current planning algorithms split an area of interest into one or several flight plans based on an estimate of the time required to complete the flight. The goal of this project is to expand this algorithm to include dynamic re-planning during flight based on several factors, including remaining battery level, area covered, wind estimation, the number of drones in the air, etc. Algorithms will first be prototyped using a simulator and then tested in the air using several drones. If successful, the algorithms will be integrated into a future release of our flight planning software.

Interested students should have knowledge in programming (C++, experience with Qt a plus), mathematical modelling and communication protocols. Experience with drones is a plus. Please send your application including a current CV and your student records (Studienerfolgsnachweis) to bernhard.rinner@aau.at.

Share

Thema: DA_NES, Diplomarbeit, Institut, NES, Studium | Kommentare (0) | Autor:

Joint master thesis with SenseFly (Lausanne)

Dienstag, 8. April 2014 7:40

As part of our SINUS research project (http://uav.aau.at), NES is offering a joint master thesis with SenseFly (http://www.sensfly.com) – a leading drone company located in Lausanne, Switzerland. Interested students are requested to contact Prof. Bernhard Rinner.

Project Description

senseFly Ltd, a spinoff of the Swiss Federal Institute of Technology (EPFL) in Lausanne, develops and commercialises mini-drones for civilian applications in the domain of 2D and 3D mapping. The maps generated with these drones are used for such applications as construction, mining logistics, forestry, conservation, crop monitoring or humanitarian aid. As part of its extensive research and development program, senseFly is looking for motivated engineers and programmers to develop, design, test and even release to market new hardware and software modules for our mini-drones. As an intern or master’s student, you will be integrated directly with our dynamic team of engineers and researchers working on new products on the cutting edge of aerial robotics. Projects have a duration of 6 months and can be tailored based on the competencies of the applicant.

The proposed thesis project involves the development of smart flight planning algorithms for swarms of fixed-wing eBee drones (https://www.sensefly.com/drones/ebee.html). Current planning algorithms split an area of interest into one or several flight plans based on an estimate of the time required to complete the flight. The goal of this project is to expand this algorithm to include dynamic re-planning during flight based on several factors, including remaining battery level, area covered, wind estimation, the number of drones in the air, etc. Algorithms will first be prototyped using a simulator and then tested in the air using several drones. If successful, the algorithms will be integrated into a future release of our flight planning software.

Interested students should have knowledge in programming (C++, experience with Qt a plus), mathematical modelling and communication protocols. Experience with drones is a plus. Please send your application including a current CV and your student records (Studienerfolgsnachweis) to bernhard.rinner@aau.at.

Share

Thema: DA_NES, Diplomarbeit, Institut, NES, Studium | Kommentare (0) | Autor:

Open Master Theses @ Smart Grid Group

Mittwoch, 26. März 2014 10:20

The Smart Grid Group of the NES-Institute is offering two new topics for Master Theses:

Occupancy Detection in Household Power Draw

Master Thesis or Research Project (Informatics, Mathematics, Information Technology):

Smart metering and accordingly, fine-grained energy demand data introduces the possibility to extract useful information from the monitored data such as appliance level demand feedback. One possible and important scope of energy consumption data is to detect the occupancy of household residents, which is fundamental to enable a various number of different smart home and home automation applications. Therefore, the task of this Master Thesis is it to identify state of the art algorithms for occupancy detection, to implement the most important and promising algorithms, to compare them and finally, to test them in a real world environment.

Knowledge: Working in a scientific way, programming language according to personal preference (Matlab, Python, C++)
Contact & Advisor: Dominik Egarter (dominik.egarter@aau.at)

Energy-Efficient LED Light Control for Smart Buildings

Master Thesis (Informatics, Mathematics, Information Technology):

The reduction of energy demand is an actual and important challenge. A big portion of today’s energy demand underlies the energy demand of household lighting systems. Therefore, the aim of this thesis is to identify an energy-efficient lighting system based on LED technology, to create and to implement a test system and to evaluate its performance compared to a todays conventional lighting system.

Knowledge: Circuit design, embedded programming (microcontroller, Raspberry Pi)
Contact & Advisor: Dominik Egarter (dominik.egarter@aau.at)

Do you have questions and are you interested for one of the topics? Do not hesitate to contact me (dominik.egarter@aau.at)!

Share

Thema: DA_Mobile Systems, DA_NES, Diplomarbeit, Institut, Lakeside Labs, NES | Kommentare (0) | Autor:

Open Topic @ Pervasive Computing: Task assignment on mobile devices

Montag, 27. Mai 2013 13:39

Finding an optimal assignment of tasks to nodes in the network is a common problem in distributed computing. Deciding which node performs what task(s) is influenced by various constraints and optimization goals (e.g. power consumption, processing load, latency …). This problem gains additional complexity when nodes are mobile, especially if they are connected via (potentially unreliable) low-power networking technologies. This master’s thesis deals with task assignment for a network of mobile nodes. In a first step, a comparative study on the state-of-the-art in task assignment should be performed. Selected task assignment mechanisms shall be evaluated for their suitability for mobile applications. Criteria like robustness to failing connections, execution speed and low communication effort should be used to compare the approaches. From the experience gained, a new approach especially for mobile networks will be designed and evaluated.

Please contact Prof. Bernhard Rinner and Univ.-Ass. Dipl.-Ing. Bernhard Dieber for details. Further open topics can be found here!

Share

Thema: DA_NES, DA_Pervasive Computing, Diplomarbeit, Forschung | Kommentare (0) | Autor:

TrustEYE: Multiple Open Theses Topics

Dienstag, 12. Februar 2013 20:59

3D Simulation Environment

In project TrustEYE at the Pervasive Computing Group we are currently offering a wide range of open topics for research projects as well as bachelor or master theses. Topics are open to students from information technology or computer science.
The overall theme of the project is the development of techniques and tools to protect privacy of monitored people in future visual sensor networks. This is especially important since such networks are used no longer only in public places but also in private environments (e.g., assisted living applications). Offered project an theses topics include:

  • design and development of virtual 3D environments (based on Half Life 2)
  • development of social gaming applications for crowdsourcing privacy evaluation
  • embedded computer vision and wireless networking on ARM-based microcontroller systems
  • security solutions for embedded systems
  • computer vision techniques for privacy protection

ARM-based TrustEYE prototype

Additional information as well as a complete list of currently open topics together with detailed descriptions and contact details can be found at the TrustEYE Open Topics site.

Share

Thema: DA_NES, DA_Pervasive Computing, Diplomarbeit, Institut, NES, Studierendenprojekt | Kommentare (0) | Autor:

Diplomarbeit: Erweiterung des Java-Frameworks FREVO

Montag, 4. Februar 2013 22:12

FREVO ist ein Open-Source Framework welches Ingenieure und Wissenschafter beim Entwurf von selbstorganisierenden Systemen unterstützt. Es ist strikt modular aufgebaut, was die Erweiterung um einzelne Funktionen für geübte Java-Programmierer einfach macht. FREVO basiert auf Vorarbeiten an der TU Wien und wurde in seiner aktuellen Form an der Alpen-Adria-Universität Klagenfurt konzipiert. Die aktuelle Version ist unter http://frevo.sourceforge.net/ verfügbar.

Anwendungen sind der evolutionäre Entwurf von Algorithmen für Multi-Agenten Systeme, wie zum Beispiel das Finden eines Spielalgorithmus für eine Roboterfussballmannschaft:

Eine kurze Einführung zum Tool findet sich im Blogartikel „Just 6 minutes of your time“.

Im Rahmen der Diplomarbeit ist ein Modul zu entwerfen welches Spiking Neural Networks als neue Representation für einen Algorithmus implementiert. Das Ergebnis ist mit Hilfe der vorgegebenen Szenarien mit den anderen bereits vorhandenen Algorithmusrepresentationen zu vergleichen.

Voraussetzungen:

  • Studium der Informatik, Informationstechnik oder Mathematik
  • Gute Java-Kenntnisse und Programmierfahrung
  • Erfahrungen im Umgang von Eclipse, SVN
  • Interesse an Neuronalen Netzwerken

Das Thema kann bei Interesse gerne aufgewertet werden, was eine wissenschaftliche Weiterführung des Themas oder ein gemeinsames Arbeiten an dem Thema gemeinsam mit einem Kollegen/einer Kollegin möglich macht (zwei DiplomandInnen oder Diplomarbeit+Softwarepraktikum).

Bei guter Leistung ist eine Unterstützung der Diplomarbeit durch ein Stipendium möglich.

Kontakt: Prof. Wilfried Elmenreich (wilfried.elmenreichaau.at) oder István Fehérvári (istvan.fehervariaau.at)

Share

Thema: DA_Mobile Systems, DA_NES, Diplomarbeit, Institut, Lakeside Labs, NES, PR_Mobile Systems, PR_NES, Praktikum | Kommentare (0) | Autor:

Stau verhindern mit neuen Technologien

Dienstag, 20. Dezember 2011 13:07

WissenschaftlerInnen arbeiten an neuen Technologien und Systemen, die automatisch erkennen können, was in ihrer Umgebung passiert. Im Herbst fand dazu eine internationale Tagung an der Alpen-Adria-Universität statt.

Wann immer Ferien oder Dienstschluss anstehen, kommt es zu großen Staus auf den Hauptverkehrsrouten. Besonders vor Baustellen und Tunnels werden die Nerven der AutofahrerInnen in kilometerlangen Staus strapaziert – eine Herausforderung für VerkehrsplanerInnen und Straßenbetreiber, die mit neuen Technologien nach Lösungen suchen.
An einer solchen Technologie wird an den beiden Instituten für Vernetzte und Eingebettete Systeme (NES) und für Informationstechnologie (ITEC) an der Alpen-Adria-Universität unter dem Projekttitel LOOK2 gemeinsam mit der ASFINAG und den Lakeside Labs geforscht. Die WissenschaftlerInnen entwickeln Systeme, mit denen für den Verkehr relevante Daten analysiert werden. Zwei solche Testabschnitte gibt es derzeit. Alle drei Minuten werden die Verkehrsdaten (Verkehrsstärke und Geschwindigkeiten), ausgewählte Umfelddaten (z. B. Temperaturen, Niederschlagsintensität) und Messdaten von intelligenten Videokameras miteinander verknüpft. Daraus entsteht ein Bild der Verkehrslage in Echtzeit, das es bereits frühzeitig ermöglicht, gegenzusteuern, sei es durch Staumeldungen über das Radio oder durch Umleitungen.

Dahinter steht ein komplexes Zusammenspielvon verschiedenen Technologien und Systemen, die erkennen können, was sich in der Umgebung der Sensoren tut. Die enorme Datenmenge, die von den Kameras geliefert wird, stellt eine besondere Herausforderung dar und „überflutet“ meist die BenutzerInnen.
Automatische Verfahren sind gefordert,die die relevanten Informationen aus den Kamerabildern extrahieren und die möglichen Konsequenzen abschätzen können.
Weltweit arbeiten WissenschaftlerInnen an der Weiterentwicklung dieser Systeme, die unterdem Begriff „Advanced Video and Signal-based Surveillance (AVSS)“ zusammengefasst
werden. Von 30. August bis 2. September 2011 trafen sich über 150 ExpertInnen auf diesem Gebiet bei der „8. IEEE Conference on Advanced Video and Signal-based Surveillance (AVSS)“ an der Alpen-Adria-Universität. Das Programm beinhaltete drei Workshops, die Hauptkonferenz und erstmals einen „Industrial Surveillance Day“.

Im Anschluss fand eine Summer School statt. Klagenfurt reiht sich in eine prominente Liste von Veranstaltungsorten, wie Boston, Genua, London oder Sydney, ein. 2012 findet die Tagung in Peking statt.

Kurzinterview mit Bernhard Rinner:

Wie gut funktionieren die Systeme auf den beiden Straßen, auf denen getestet
wird, derzeit?
Unsere Prototypen klassifizieren den Verkehrszustand am Testabschnitt in Echtzeit. Die ersten Auswertungen sind bereits sehr vielversprechend. Verbesserungen
erwarten wir uns durch die Kombination mehrerer Verkehrssensoren.

Die Technologien und Systeme sollen erkennen, was sich in der Umgebung der Sensoren tut. Werden sie hinkünftig auch „auffälliges“ Fahrverhalten von AutofahrerInnen erkennen?
Das hängt naturgemäß davon ab, was man unter „auffällig“ versteht. Viele Forschungsprojekte beschäftigen sich derzeit mit „lernenden“ Systemen,die ungewöhnliches Verhalten selbst erkennen.

In welchen Bereichen kommt diese Technik noch zum Einsatz?
Die Kenntnis über den aktuellen Zustand der Umgebung ist für viele Anwendungen sehrwichtig. Denken Sie nur an Fahrerassistenzsysteme, an die Robotik, oder an „smart
homes“. Hier sind die notwendigen Computer und Sensoren im Gesamtsystem „eingebettet“und müssen selbst erkennen, was gerade im Umfeld passiert, um richtig reagieren zu
können. Wir sehen viele Alltagsanwendungen – ein wichtiger Aspekt dabei ist der Schutz der Privatsphäre.

(aus Unisono 411, S13)

Share

Thema: DA_NES, DA_Pervasive Computing, Diplomarbeit, Forschung, Medienpräsenz, Unisono | Kommentare (0) | Autor: