Long periods of idle sitting are known to significantly increasing the risk for overweight and obesity which, in turn, dramatically increase the risk of several debilitating, and deadly diseases, including diabetes as well as heart disease. Despite the negative effects of long sitting periods, the number of hours that people remain continuously in a state of low physical activity keeps increasing in most first world countries, including Germany. This not only causes issues for affected individuals, e.g. by reducing their quality of life as well as their overall life expectancy, but it also increases the stress on the health care and the pension system as well as the labor market.

Economists, social scientists, and psychologists have shown that comparatively inexpensive nudges can effectively influence the behavior of people without reducing their freedom of choice. The goal of this project group is to develop a mobile app with an associated set of backend services that applies the idea of nudging to reduce the periods of physical inactivity of university members. Towards this end, the resulting platform shall not only provide the means for context-aware nudges, but it shall also support the systematic analysis of their effectiveness for different users.

From a theoretical perspective, the project group will cover concepts related to various forms of nudging (e.g. notifications, gamification, etc.) as well as the development of context-aware mobile applications (e.g. signal processing, machine learning, etc.).  From a practical perspective, the project group will cover full stack development with the Jakarta EE framework (Java), Android (Kotlin), and iOS (Swift). Students taking this course must be fluent in at least one object-oriented programming language and should be able to quickly apply their knowledge to other languages exhibiting similar abstractions.

In this project, we want to scan environments as 3-dimensional point clouds using a depth scanner mounted on a drone. The environment will be rendered live to a viewer.

In a typical use case, the drone will be placed in an unknown environment and should gradually map the area, autonomously navigating. Hereby, point cloud registration is performed on a server, to which the drone streams the sensor data captured in flight. The server then reconstructs the 3-dimensional area using a SLAM algorithm and sets waypoints for the drone to further explore and capture the environment.

Drone control is abstracted away through an autopilot, which allows navigation through high-level commands. Obstacles in the path of the drone should be automatically detected and avoided by the drone.

Cross-Chain Brücken ermöglichen die Interoperabilität zwischen dezentralen Blockchains und Netzwerken und werden mit der Vielzahl an verschiedenen Systemen zunehmend wichtiger. Sie ermöglichen den Datenaustausch zwischen ansonsten unabhängigen Blockchain Netzwerken (z.B. Bitcoin, Ethereum, Solana, ...), indem die Brücke auf verschiedenen Blockchains "mithört" und die Daten dann in einem eigenen Netzwerk weiterverteilt. Mit einem derzeitigen Marktvolumen von über 6 Milliarden US Dollar sind sie eine wichtige Stütze für Zukunftstechnologien und dezentrale Märkte. Jedoch kam es in der Vergangenheit immer wieder zu Verlusten durch Schwachstellen in den Brücken.

Im Rahmen der PG lernen Studierende Cross-Chain Brücken kennen. Sie lernen dabei die Schwachstellen, die dabei auftreten können, zu finden. Außerdem werden Kenntnisse in der Programmiersprache Solidity und im Generellen zu Blockchain Netzwerken vermittelt.

Während der PG soll eine funktionierende Cross-Chain Brücke zwischen EVM-basierten Blockchains (z.B. Ethereum) aufgebaut werden. Ziel der PG ist das Design und die Implementation einer sicheren, dezentralen Cross-Chain Brücke.

Modern software development approaches, such as DevOps, facilitate continuous software integration and delivery (CI/CD). Such continuous CI/CD can be supported by Just-in-time (JIT) defect prediction techniques. These techniques provide feedback on whether a code change committed to the software repository is likely to contain defects. This immediate feedback allows practitioners to make timely decisions regarding potential defects.  However, a prediction model may deliver false predictions, that may negatively affect practitioners' decisions. False positive predictions lead to unnecessarily spending resources on investigating clean code changes, while false negative predictions may result in overlooking defective changes.  Knowing how uncertain a defect prediction is, would help practitioners to avoid wrong decisions.

A potential solution to address the problem of quantifying prediction uncertainty is conformal prediction (CP). CP can be combined with any prediction model that provides some heuristic notion of uncertainty, such as prediction probabilities. CP uses a small amount of additional calibration data to convert the heuristic notion of uncertainty into a rigorous one. Instead of generating an output in the form of a single label, CP generates prediction sets that are guaranteed, with probability 1-α, to contain the true label. In the optimal case, the prediction sets consist of a single label.

The goal of this project group is to apply and systematically evaluate conformal prediction (CP) as a rigorous uncertainty quantification approach on top of state-of-the-art JIT defect predictors. The students will first review the literature on JIT defect prediction and select suitable state-of-the-art approaches to work with. They will then apply CP on the selected JIT approach, using real-world large-scale defect datasets (e.g., such as from the OpenStack and Apache open-source projects). Provided with access to a modern GPU-Server (equipped with four NVIDIA GeForce RTX 4090 GPUs), they will perform experiments to analyze: (1) how often CP can provide guarantees for JIT defect predictions; and (2) how many false JIT defect predictions CP can filter out. Based on an in-depth analysis of the experiment results, the students will discuss potential directions for enhancing the performance of conformal JIT defect prediction.