IDP · AI-assisted prototype for the logistics sector
Interdisciplinary Project · Master Informatics · ~1 semester · equivalent formats welcome
Prototype a decision layer for logistics teams that turns messy operational inputs into structured, explainable planning suggestions.
ARCIS builds applied-AI decision systems for the logistics sector — the people behind trucks, loads, schedules and the daily flow of goods. This project focuses on prototyping a layer that helps operators make faster, more structured decisions on top of the tools they already use. Expected outcome: a focused prototype and technical documentation.
Project scope
- Extracting structured data from operational inputs (spreadsheets, documents, emails)
- Validation logic for missing or noisy fields
- Matching / assignment under real-world logistics constraints
- Optimization methods (assignment, heuristics, constraint programming)
- Explainable outputs and human-in-the-loop review
Research & engineering questions
- How can unstructured inputs be converted into reliable planning data?
- How can assignment be modeled under incomplete information?
- How can optimization results be explained to non-technical operators?
Ideal profile
Target program: TUM Master of Science in Informatics (M.Sc. Informatics). The role is framed as a TUM Interdisciplinary Project (IDP). Curiosity for the logistics sector and applied AI is essential; prior logistics knowledge is not required. Working-student or internship arrangements for equivalent technical work are also welcome.
- Python · data processing
- LLM-based extraction
- OR-Tools · optimization · algorithms
- TypeScript / React (optional)
What you get
- You will earn 16 ECTS credits at TUM.
- Exclusive access to TUM Incubator, entrepreneurial coaching, and workshops.
- Network with industry leaders, founders, and senior professionals.
Key facts
- 6 months part-time (approx. 20 hours/week) or alternatively, 3 months full-time.
- Ideally 2–5 Master’s students, but individual applications are also welcome.
- Project can start immediately or upon agreement.
Application
Please send:
- CV
- Transcript of records
- Short motivation note
- Optional: GitHub, portfolio or previous technical work