Masters Projects

At the Control Theory and Systems Biology Laboratory, led by Prof. Dr. Mustafa Khammash, we are developing advanced automation solutions for biological laboratory operations. Liquid handling robots are crucial tools to automate repetitive, time-consuming tasks. By automating these processes, scientists can focus on research while benefiting from improved experimental reproducibility and the ability to conduct more sophisticated experiments. However, state-of-the-art liquid handling systems need to be explicitly programmed for each task, and run, for the most part, in open loop. This means that in many cases once a protocol is started, the robot follows pre-programmed movements without any real-time feedback about the success of its operations. This blind execution can lead to undetected errors such as incorrect liquid volumes, or failed transfers. Furthermore, adapting protocols to different labware or unexpected situations requires manual reprogramming, making these systems unflexible.

Computer vision can address some of these limitations by providing real-time feedback about the robot's operations, enabling automatic error detection and correction, and allowing for more adaptive and intelligent liquid handling workflows.

## Tasks In this ambitious and open-ended project the student will have the opportunity to go through a diverse set of tasks and set up a computer vision system for a liquid handling robot, namely: ### Setting up: - Hardware integration of cameras operating in the visible and thermal infrared ranges into the robot (Opentrons OT-2) - Software integration of existing cameras in the robot frame - Review of state-of-the-art computer vision applications in liquid handling

### Computer Vision: - Liquid level detection in tips and labware - Calibration of measurements, error estimation - Labware detection, identification and pose estimation - 3D reconstruction of the robot workspace - Error detection and classification

### Software integration - Integration of computer vision with the robot control software

### Deliverables - Written report - 20 minute presentation - Structured and well-documented code

The tasks listed are suggestions and the student is encouraged to propose and implement their own ideas.

## Possible Extensions - Contribute upstream to open source projects, e.g. to Opentrons source code

## Requirements Candidates should: - Be creative, independent, and self-motivated - Have strong programming skills, particularly in Python - Have experience with computer vision libraries (OpenCV, etc.) or the willingness to learn them - Basic understanding of machine learning concepts - Interest in automation and robotics - Willingness to work in a laboratory environment (BSL1) - Interest or experience in biology is a plus - Experience with ROS or similar robotics frameworks is a plus

The project earliest starting date is September 2025. It will remain open until a suitable candidate is found.

The ultimate goal of the project is to develop a robust computer vision system that can reliably monitor liquid handling operations, detect potential errors, and validate successful liquid transfers in real-time. The output of the project will be a software package that can be integrated into existing liquid handling robots, complementing its existing functionality.

For any questions regarding this posting please contact: Dr. Samuel Balula (samuel.balula@bsse.ethz.ch)

In your application please include: - A brief (1-2 paragraphs) cover letter with your reasoning to apply to this project - Curriculum Vitae - Transcripts

# Location The Control Theory and Systems Biology Laboratory is located in the new D-BSSE building in Basel. The student is expected to come in person to the lab to conduct experimental work whenever that is necessary. There's flexibility for online meetings and remote work.

At the Control Theory and Systems Biology Laboratory, led by Prof. Dr. Mustafa Khammash, we are developing technology to enable a seamless interface between the biological and digital worlds. This enables the study of synthetic biological circuits, prototyping mixed controllers (cybergenetics), and ultimately the development of synthetic control systems implemented exclusively with biological parts, with a wide range of applications in medicine and biotechnology.

A key aspect of this interface is the measurement of state information from biological samples. We are particularly interested in obtaining high-frequency data (in the context of biological timescales), with high throughput and quality, from samples in culture under controlled conditions.

Optical measurements are versatile and widely useful for measuring a variety of parameters, for example, the concentration of cells, pH, or individual gene expression levels. This project focuses on extracting information from spectroscopy data using an interpretable data-driven method, and builds upon preliminary work that has been done in the lab, with a custom built, low-cost fluorescence spectrometer.

In this ambitious project, the student will have the opportunity to creatively apply or create robust and interpretable methods to extract information from spectroscopy data.

## Tasks - Mathematical modelling of scattering, fluorescence and absorption in the context of spectroscopy - Data analysis, model selection, and fitting - Design of experiments - Automated interpretation of results from known spectra - Statistical guarantees on performance - Documentation and presentation of results

## Possible Extensions - User interface for the method and integration with automated experimental setups - Publication of results in a peer-reviewed journal or conference proceedings

## Candidates should - Be creative, independent, and self-motivated - Be familiar with system identification methods, optimization, and statistics - Be willing to perform experimental work in a biology lab (BSL1) with assistance from lab members - Have at least basic knowledge of optics, control theory and machine learning - Have strong programming skills. The project will be implemented preferably in Python. - Knowledge of biology is a plus but not required

### Deliverables - Written report - 20 minute presentation - Structured and well-documented code

The project's earliest starting date is September 2025. It will remain open until a suitable candidate is found.

The primary goal is to develop and validate a comprehensive spectral unmixing framework that can automatically extract meaningful biological parameters from complex spectroscopy data.

The successful completion of this project will significantly enhance the laboratory's capability to perform high-throughput, real-time biological measurements, advancing research in synthetic biology and cybergenetics.

For any questions regarding this posting please contact: Dr. Samuel Balula (samuel.balula@bsse.ethz.ch)

In your application please include: - A brief (1-2 paragraphs) cover letter with your reasoning for applying to this project - Curriculum Vitae - Transcripts

# Location The Control Theory and Systems Biology Laboratory is located in the new D-BSSE building in Basel. The student is expected to come in person to the lab to conduct experimental work whenever necessary. There is flexibility for online meetings and remote work.

At the Control Theory and Systems Biology Laboratory, led by Prof. Dr. Mustafa Khammash, we are developing advanced automation solutions for synthetic biology research. Automated bioreactor systems are essential tools for conducting reproducible, high-throughput experiments in controlled environments, enabling precise manipulation and monitoring of biological systems over extended periods of time.

Current automated bioreactor platforms offer limited scalability and often lack integrated control capabilities for advanced synthetic biology applications. While existing systems provide basic turbidostat functionality, they typically lack accurate measurements of biologically relevant parameters. This is a bottleneck for cybergenetic experiments (control systems for living cells)

This project focuses on developing a software framework for a high-throughput turbidostat platform built with an automatic liquid handling robot, enabling real-time closed-loop control of multiple variables simultaneously.

## Tasks In this project, the student will have the opportunity to develop a complete automated bioreactor control system, namely:

• Design and implement control systems for a bio reactor various variables
• Create user-friendly interfaces for experiment design and monitoring
• Build comprehensive data logging and storage systems enabling traceability and failure recovery
• Develop real-time visualization and monitoring
• Interface with sensors and actuators via network protocols, including the liquid handling robot

### Deliverables - Written report - 20 minute presentation - Structured and well-documented code

### Possible Extensions - Priority scheduling of tasks for the liquid handling robot - Integration with laboratory information management systems (LIMS) - Publication of results in a peer-reviewed journal or conference proceedings

### Requirements Candidates should:

• Have strong programming skills
• Be familiar with control theory and interested in automation
  • Be interested in interdisciplinary work combining engineering and biology
  • Experience with Linux, ROS or similar robotics frameworks or the willingness to learn them
  • Willingness to work in a laboratory environment (BSL1)
• Basic understanding of biology and synthetic biology concepts is a plus but not required
  • Experience with microcontrollers and embedded systems is a plus but not required

The project's earliest starting date is September 2025. It will remain open until a suitable candidate is found.

The primary goal is to develop and validate a comprehensive software framework for a high-throughput automated turbidostat platform that enables precise, real-time control of multiple biological cultures.

The successful completion of this project will significantly advance the laboratory's capabilities for conducting sophisticated, automated synthetic biology experiments, enabling new research directions in cybergenetics and synthetic biological control systems.

For any questions regarding this posting please contact: Dr. Samuel Balula (samuel.balula@bsse.ethz.ch)

In your application please include:

• A brief (1-2 paragraphs) cover letter with your reasoning for applying to this project
• Curriculum Vitae
• Transcripts
• Examples of relevant programming projects or technical work (if available)

# Location

The Control Theory and Systems Biology Laboratory is located in the new D-BSSE building in Basel. The student is expected to come in person to the lab to conduct experimental work and system integration whenever necessary. There is flexibility for online meetings and remote software development work.

The project is designed for a Master thesis (MA), but we are open to adjust it to other project types (SA, BA) for outstanding candidates.

At the Control Theory and Systems Biology Laboratory, led by Prof. Dr. Mustafa Khammash, we are developing advanced automation solutions for biological laboratory operations. Liquid handling robots are crucial tools to automate repetitive, time-consuming tasks. By automating these processes, scientists can focus on research while benefiting from improved experimental reproducibility and the ability to conduct more sophisticated experiments.

However, state-of-the-art liquid handling systems - which are at its core 3 axis gantry systems - typically use simple point-to-point movements with predefined speeds, resulting in inefficient trajectories and longer execution times. Additionally, the placement of labware and instruments on the deck is usually not optimized, leading to unnecessary long paths and increased execution times.

Motion optimization and intelligent trajectory planning can address these limitations by generating efficient, smooth paths while considering the physical constraints of liquid handling robots. This includes optimizing for shorter paths, varying speeds based on payload and movement type, and suggesting optimal deck layouts for specific protocols.

## Tasks In this ambitious and open-ended project the student will have the opportunity to go through a diverse set of tasks and develop a motion optimization system for a liquid handling robot, namely:

### Setting up: - Review of state-of-the-art trajectory optimization methods for robotic systems - Understanding the robot's hardware and software architecture

### Modeling and Simulation: - Development of a 3D kinematic model of the robot (Opentrons OT-2) - Implementation of basic simulation of the robot for testing

### Motion Planning and Optimization: - 3D model of the robot workspace (physical constraints) - Development and testing of different trajectory optimization algorithms considering: * Path length and execution time * Biological requirements * Safety and collision avoidance - Deck layout optimization

### Software Development: - Integration with robot control software - Development of visualization tools for protocols

### Deliverables - Written report - 20 minute presentation - Structured and well-documented code

The tasks are open-ended and the student is encouraged to propose and implement their own ideas.

### Possible Extensions - Contribute to open source projects - Dynamic reconfiguration of the robot's workspace

### Requirements Candidates should:

- Have strong programming skills, particularly in Python - Have experience with optimization algorithms or the willingness to learn them - Basic understanding of robotics and motion planning concepts - Interest in automation and robotics - Willingness to work in a laboratory environment (Biosafety level 1) - Interest or experience in biology is a plus - Experience with ROS or similar robotics frameworks is a plus

The project earliest starting date is September 2025. It will remain open until a suitable candidate is found.

The ultimate goal of the project is to develop a motion optimization system that can generate efficient, safe trajectories for liquid handling operations while considering the physical constraints of the system and the nature of the payload. The output will include a 3D model of the robot, optimization algorithms for trajectory generation, and tools for deck layout optimization.

For any questions regarding this posting please contact: Dr. Samuel Balula (samuel.balula@bsse.ethz.ch)

In your application please include: - A brief (1-2 paragraphs) cover letter with your reasoning to apply to this project - Curriculum Vitae - Transcripts

# Location The Control Theory and Systems Biology Laboratory is located in the new D-BSSE building in Basel. The student is expected to come in person to the lab to conduct experimental work whenever that is necessary. There's flexibility for online meetings and remote work.