Internship (m/f/d) - Building a Motorized Goniometric Light-Scattering Setup with Machine-Learning Image Analysis

16.10.2026, Studentische Hilfskräfte, Praktikantenstellen, Studienarbeiten

When light passes through a turbid sample such as a bioprocess medium, it is both scattered and absorbed. These two effects overlap in the measured signal, which makes optical measurements harder to interpret. This is a major obstacle for optical process monitoring, for example in the production of biopolymers, green fuels, and pharmaceuticals. Existing solutions that separate the two effects are expensive or require complex optical setups.
Our Lasers and Particles research Group at the Chair for Analytical Chemistry and Water Chemistry at TUM monitors bioprocess media with optical spectroscopy and develops its own measurement devices, which we either 3D-print in our lab or build in the mechanical workshop. Thus, we cover the whole chain from idea through design and manufacturing to measurement and data analysis.

In this project, we are developing a new, low-cost approach that separates scattering and absorption using camera images of the light distribution in a sample. We evaluate these images with physics-based models (Mie theory) and data-driven models such as Partial Least Squares regression and neural networks (NNs). A first prototype has already produced promising results. The next step is an automated measurement platform, and this is where you come in.

Your task in this project:
• Design the setup in a CAD software and 3D print it
• Installing and controlling of the motorized moving parts in multiple dimensions
• Perform automated measurement series with scattering and absorbing samples
• Evaluate the measured images using different image processing models
• Compare the measurements with a simulated digital twin
• Test and improve model generalizability
• Your own ideas

What you bring:
• Programming experience, ideally in Python and/or R
• Interest in hardware-related work (3D printing, microcontrollers, motor control) and image and data analysis
• No prior knowledge of chemistry or optics required, just curiosity about interdisciplinary work

What you gain:
• Hands-on experience at the interface of computer science, engineering, and natural sciences
• A project with a real application in bioprocess monitoring
• Supervision by and insight into a wildly interdisciplinary team

Start:
As soon as possible.

Location:
Our team is located at Georg-Heberer-Straße 11 in Großhadern, Munich. Office space is available, and for the hardware part, your presence is required. For the programming part, home office may be more convenient for you.

Corresponding lectures:

Current Research in the Instrumental Analysis of Trace Components 1 (CH3122/NAT0170)
Analytische Chemie (CH0107, 2 SWS)
Fortgeschrittene analytische Verfahren (CH4115, 4 SWS)

Organisation:
Simply contact Florian Gollinger (florian.gollinger@tum.de) via email.

Kontakt: florian.gollinger@tum.de