Open Theses

We are pleased that you are interested in a thesis at the Department of Computer Science with a focus on Embedded Systems. Please inform yourself in advance which topic is suitable for you. Afterwards, please make an appointment for the first meeting with the respective supervisor by e-mail.

The make  an appointment in advance is not only a sign of good organization, but also a matter of courtesy: as a rule, the supervisors are rather reluctant when you show up in the office unannounced.

Your first contact email should include:

  • A meaningful subject.
  • Your request (Bachelor thesis / master thesis).
  • How you have become aware of the topic and your motivation to edit the topic (just about 2-3 sentences).
  • The completed questionnaire for prospective students for theses.
  • The request for an appointment for a personal conversation.

Gladly seen but not obligatory:

  • Transcript of Records
  • Overview of past projects

 

 

Visualization of Biofeedback Data using a Smart-Watches

Supervisor:  Dr. Gerold Hölzl

Student: -

Biofeedback is an established technique to gain greater awareness of many psychological functions primarily using instruments that provide information on the activity of those same systems, with the goal of being able to manipulate them at will.

A current ongoing research project investigates in the question how a Smart-Watch based biofeedback training system using HR, HRV and a derived Stress Indicator can be used by people at home to achieve a noticeable, long-term benefit from the training.

The Goal of this thesis is to investigate in creative approaches to visually represent the measurements on the watch face to quickly inform the user but not people who may spy on the device.


The developed Concepts have to be implemented on an Android based Smartwatch and evaluated using the departments Biofeedback Framework in a pilot study.

Requirements: Successful candidates will have a strong background in computer science, mobile and embedded systems, or comparable study programmes, and a high interest in playing with sensor and web technology and knowledge in distributed client/server based systems.

Therapy Support App (System Development)

Supervisor: Prof. Dr. Matthias Kranz

Student: -

Mental health and psychological problems are of high importance to modern society[1]. Within 20 years, the amount of persons treated stationary in hospitals raised from about 770.000 persons in 1994 to 1.238.000 persons in 2014[2]. In addition to immense personal suffering, the costs for the health system and for the economy are exploding. Contributing to an improved treatment of patients is therefore an important goal.

In many forms of therapy, the patient is required to adhere to a medication plan, log blood values (such as insulin levels), alcohol and drug usage, keep track of his sleep and life habits and keep notes on his personal well-being.

In the course of this thesis, a smartphone application is to be developed, providing an appealing, pleasuring and effective user interface for the patient.

The application shall feature the tracking of sleep times, medicine intake according to a definable scheme and plan (doses, times, frequencies, ...), a mood diary and a way note current life behaviours (partying). The application shall be allowing easy creating of medication plans, e.g. by scanning the barcode of the medication using the smartphone's camera. Reminders on medication shall be given which shall be confirmed by the user, if possible. The application shall feature the creation of reports, e.g. on medicine intake and the export of the data in standard formats such as CSV or JSON. It should, by a calendar function, also allow the user to administrate his/her therapy appointments.

The goal is to create an application that is both useful and usable, and goes beyond the currently available applications. It should differentiate itself by its clarity, simplicity and utility against similar apps in the App Store. The app shall motivate the user to use it on a long-term basis.

In this step, no interactions or contact with real patients will be required.

Experts from the medical and psychological domain are at any time co-supervising and supporting this thesis with their knowledge and feedback.

Requirements: Successful candidates will have a strong background in computer science, computer engineering, or comparable study programmes, preferable a deep knowledge in Android programming, Java, human-computer interaction, sketching, interaction design, or user experience design.

Stress and Concentration Detection (System Development)

Supervisor: Prof. Dr. Matthias Kranz

Student: -

Stress affects the performance of users in a negative way. Even the most trained and competent users make more errors when exposed to stress.

In this thesis, an existing platform consisting of an eye tracker and an extended software development environment shall be used and complemented with additional sensors. The current system allows attention focus estimation based on the eye tracking data and to correlate this information with given lines and fragments of source code. The system is intended to estimate which parts of source code are potentially more error-prone than others. It shall complement other approaches to software quality estimation.

The Myo Band offers EMG sensors and an API for accessing the data. In this thesis, the Myo band shall be connected to the present system. It shall be investigated if and to which extent EMG data collected while writing source code can be used to make statements on software quality.

Experts from the medical and psychological domain are co-supervising and supporting this thesis with their knowledge and feedback.

Requirements: Successful candidates will have a strong background in computer science, computer engineering, or comparable study programmes, preferable a deep knowledge in Android programming, Java, human-computer interaction, sketching, interaction design, or user experience design.

 

 

Therapy Support via Wearable and App (System Development)

Supervisor: Prof. Dr. Matthias Kranz

Student: -

Mental health and psychological problems are of high importance to modern society[1]. Within 20 years, the amount of persons treated stationary in hospitals raised from about 770.000 persons in 1994 to 1.238.000 persons in 2014[2]. In addition to immense personal suffering, the costs for the health system and for the economy are exploding. Contributing to an improved treatment of patients is therefore an important goal. In many forms of therapy, the patient is required to adhere to a medication plan, log blood values (such as insulin levels), alcohol and drug usage, keep track of his sleep and life habits and keep notes on his personal well-being.

In the course of this thesis, a smartphone application and a smartwatch application is to be developed, providing an appealing, pleasuring and effective user interface for the patient.

Smartphone Application: The application shall feature the tracking of sleep times, medicine intake according to a definable scheme and plan (doses, times, frequencies, ...), a mood diary and a way note current life behaviours (partying). The application shall be allowing easy creating of medication plans, e.g. by scanning the barcode of the medication using the smartphone's camera. Reminders on medication shall be given which shall be confirmed by the user, if possible. The application shall feature the creation of reports, e.g. on medicine intake and the export of the data in standard formats such as CSV or JSON. It should, by a calendar function, also allow the user to administrate his/her therapy appointments. The application shall track interaction and activity to infer the mental state of the user.

Wearable Application: The application shall acquire bio signals from the user, such as sleep times, pulse, heart rate, conductive skin response or other data. The data shall be logged and analysed with the goal of recognizing the mental condition of the user. This data shall complement data from the smartphone.

The goal is to create an application that is both useful and usable, and goes beyond the currently available applications. It should differentiate itself by its clarity, simplicity and utility against similar apps in the App Store.

In this step, no interactions or contact with real patients will be required. Experts from the medical and psychological domain are at any time co-supervising and supporting this thesis with their knowledge and feedback.

Requirements: Successful candidates will have a strong background in computer science, computer engineering, or comparable study programmes, preferable a deep knowledge in Android programming, Java, human-computer interaction, sketching, interaction design, or user experience design.

 

Neurofeedback for Therapy (App Development)

Supervisor: Prof. Dr. Matthias Kranz

Student: -

Mental health and psychological problems are of high importance to modern society[1]. Within 20 years, the amount of persons treated stationary in hospitals raised from about 770.000 persons in 1994 to 1.238.000 persons in 2014[2]. In addition to immense personal suffering, the costs for the health system and for the economy are exploding. Contributing to an improved treatment of patients is therefore an important goal. The thesis shall focus on selected mental and psychological issues such as panic attacks.

Neurofeedback is concern with measuring biological parameters such as pulse, galvanic skin response or other bodily parameters and making these parameters available for interpretation to the human. An example could be to measure the heart rate, visualize it and support the user in calming himself down. Neurofeedback is a proven approach in the treatment of many psychological disorders.

Mobile neurofeedback is concerned with providing the user with neurofeedback outside a stationary therapy setup. The idea is to complement classic therapeutical treatment with individual access to biological values to allow the user for additional learning and deepening of the acquired competences. It would be interesting to investigate if the learning rate can be increased, if more exercises are down with the patient having now individual options, or if long-term impact (post-therapy) can be increased or stabilized.

In the thesis, an existing special wearable device and/or a standard prosumer wearable smartwatch shall be used for measuring biological values. The data and platform is accessible via an existing API. The data is to be transferred via Bluetooth to an Android-based smartphone. The current Android Nougat shall be used as target environment, with compatibility enabled for Android Marshmallow. The app shall implement the process of biofeedback training on the smartphone. For this, a basic app showing biofeedback and instructions as audio/video need to be implemented. Logging of interactions shall be used to complement self-reporting of the training.

In this step, no interactions or contact with real patients will be required.

Experts from the medical and psychological domain are at any time co-supervising and supporting this thesis with their knowledge and feedback.

Requirements: Successful candidates will have a strong background in computer science, computer engineering, or comparable study programmes, preferable a deep knowledge in Android programming, Java, human-computer interaction, sketching, interaction design, or user experience design.

 

Web Based Real Time Data Sampling (System Development)

Supervisor:  Dr. Gerold Hölzl

Student: -

The analysis and interpretation of heterogeneous sensor data by using mathematical, statistical and pattern classification methodologies is a late-breaking research topic.

The collection of the sensor data itself is the most time consuming part of the workflow. The data has to be recorded and must be labelled to be analysed afterwards.

Goal of this work is the development of a Web based Framework that enables users to independently collect sensor data, to record a synchronized video as ground-truth, to label the collected sensor data, and finally to send the data to a server for later analysis.

Requirements: Successful candidates will have a strong background in computer science, mobile and embedded systems, or comparable study programmes, and a high interest in playing with sensor and web technology and knowledge in distributed client/server based systems.

Real-time Industrial Information Source Map (System Development)

Supervisor: Dr. Gerold Hölzl

Student: -

Being on the edge of Industry 4.0 with its vision of computerized manufacturing, many companies face the problem of not knowing which digital information is present and collectable in their factories.

To support them in their transition process to smart manufacturing, a tool to build a virtual plan of the factory with annotated information in terms of the information that can be captured from the sketched devices is needed.

Goal of this work is to define a Framework to build a digital information map with the possibility of integrating real time captured data from the sketched smart machines.

Requirements: Students must be interested in APP development, visualization and abstraction of sensor data, and modelling machine parameters for rule based notifications.

NFC and Sensor Fusion for Indoor Positioning

Supervisor: Dr. Gerold Hölzl

Student: -

Indoor Positioning is still a crucial issue for the success of a lot of smart home systems und additionally in smart manufacturing processes. Based on the identified location of people in households, setups of devices like lightning, heating, etc. can be performed automatically by the system.

Industrial available systems (like UbiSense) are extremely expensive and the effort to install them makes them unusable for home appliance.

Most algorithms are based on using the received signal strength indicators as a measure for the position, so users have to build a map for locations they want to identify. This is a very time consuming process.

NFC (Near Field Communication) is a new technology and a set of protocols that enable electronic devices to establish radio communication with each other by touching the devices together, or bringing them into proximity to a distance.

The Idea of the project is to use NFC enabled devices to build up a smart indoor positioning systems that can detect relevant positions of people, act accordingly, and is easily extendable and adaptable to the used context.

The goal of this project is to:

  • evaluate pros/cons of using NFC as indoor positioning technology.
  • compare different NFC enabled technologies: beacons vs. smartphones vs. wrist-worn smartwatches.
  • investigate in the achievable accuracy and runtime with NFC as indoor positioning technology.
  • Estimate the effort to setup NFC enabled indoor positioning systems.
  • Develop a demonstration setup and framework as a proof of concept for using NFC enabled technology to infer users' location as input data for smart home appliances.

Requirements: Students must be interested in APP development, visualization and abstraction of sensor data, and modelling machine parameters for rule based notifications.

Positioning hints while using smart glasses

Supervisor: Lukas Witzani

Student: -

When using smart glasses, the readability of their displays often suffers from the strong brightness of the background that lies in the line of sight of the user (eyes – display – background). However, darker backgrounds could possibly improve the display quality in terms of readability. How could modern smart glasses’ on-board technology help to ensure the best possible display quality at all times?

In this thesis, the field-of-view (FOV) of modern smart glasses such as Epson Moverio BT-300 should be examined.

Specifically, you should:

  • Find a way to access the smart glasses’ camera in an own Android app
  • Compare the FOV of the camera with the FOV of the user’s eyes through the smart glasses’ display
  • Analyze the camera’s pictures and search for darker areas
  • Create an algorithm that utilizes above information to help the user find the best possible head position in terms of reading experience, meaning that the smart glasses’ display is in front of a nearby dark background
  • Think of and implement possible types of display modes using the developed algorithm, e.g.:
    • showing the display only if user looks directly at the darker background (giving visual hints in which user has to look)
    • showing the display at all times while giving visual hints where to look in order to improve the quality of the display (see picture above)

The algorithm can be tested in the domain of text entry on smart glasses using a smartphone for keyboard input (system already exists). Preferably the enhanced system is evaluated by running a user study.

Requirements: Successful candidates should have a strong background in computer science, computer engineering or comparable study programs, initial experience with graphic libraries such as OpenGL ES and preferable knowledge in Java and mobile application programming (Android).

Improving Text-Entry in Virtual Reality

Supervisor: Lukas Witzani

Student: -

Alphanumeric text entry is a challenge for Virtual Reality (VR) applications. Existing consumer Virtual Reality systems, such as HTC Vive, Oculus Rift, or Samsung’s Gear VR use methods such as hand held controller, head or gaze direction, or a game controller. Such methods are tedious and slow and usually are used to enter very short texts, such as passwords and names.
In preceding work a prototype for a virtual reality text entry system has been developed. The system currently uses standard text input devices (physical keyboard, touchscreen keyboard) and transfers user’s experience with these kind of devices into virtual reality. Beside virtual keyboard representations of both kinds, users are able to see their fingertips. At the moment an Optitrack camera system is used for tracking the position/rotation of all keyboards, the user’s head and the user’s finger-tips.

There are different possibilities for thesis-foci which address the improvement of the current prototype to increase typing speed and reduce error rates inside VR:

  • increase usability by investigating/integrating other tracking technologies (e.g. leap motion)
  • investigate possibilities (e.g. inverse kinematic) to show virtual representations of whole hands and not only user’s fingertips
  • investigate the use of depth pictures (e.g. through Microsoft’s Kinect) to show user’s real hands and keyboard inside VR
  • investigate/integrate the possibility to use an auto correction system (e.g. Bing’s spell checker)
  • generally improve the system (e.g. use greater touch screen, display bigger virtual representations, rethink highlighting of buttons, ... )

Changes on the system should be evaluated by running a user study.

Requirements: Successful candidates should have a strong background in computer science, computer engineering or comparable study programs, preferable experience with Unity3D and knowledge in Java, C# and mobile application programming (Android).