TU Munich: Letter of motivation
Max 2 A4 pages
Explain, in no more than two A4 pages, why you have chosen this degree program and TU Munich, and what abilities, talents, interests, and ambitions make you particularly suited to it. For German-taught programs this is normally written in German.
What it’s really asking
TUM wants three linked things: why this specific program, why TUM, and concrete evidence that you can succeed in it. It is a focused case for fit with one degree, not a personal narrative, and it doubles as the basis for your interview if you land in the borderline band.
Why they ask it
The letter only appears for programs with an aptitude assessment, and it carries weight exactly when your grades alone are not decisive. A clear, evidence-led letter can move a middle-band applicant toward admission, and a vague one wastes the one chance you have to speak for yourself.
Three ways in
Mine the module handbook
Open the module handbook for the program and pick the two or three first-year courses that connect to something you have actually done, then build the letter around those links.
Lead with evidence
List your strongest pieces of evidence first (a project, an internship, a research task, a competition) and write a sentence of result or insight for each before you draft prose.
Write for the interview
Decide, for every claim, how you would answer it in a 20-minute interview; if you cannot defend it out loud, leave it out.
✕ Weak opening
“Ever since I was a child, I have been fascinated by technology and dreamed of studying at a world-class university in Germany.”
✓ Strong opening
“Rebuilding a failed water-level sensor three times taught me why signal noise, not the code, was the real problem, and it is why TUM's measurement and control modules pulled me to this program.”
Dear Admissions Committee,
I am applying for the M.Sc. in Robotics, Cognition, Intelligence because I want to build robots that handle the soft, deformable, unpredictable objects that current systems still drop, and TUM is one of the few places where I can do this across the exact three disciplines the program names: mechatronics, machine learning, and computational neuroscience.1My interest is not abstract. During my third undergraduate year I led the manipulation subteam of our RoboCup@Home entry. Our service robot could navigate a mock apartment and recognize 40 household items, but it failed the grasping task in 6 of 10 trials. The failures were not random: it crushed a paper cup and let a folded towel slip. I spent two months instrumenting the gripper with FSR pressure sensors and logging every grasp.2The data showed our controller used a single force threshold for every object, so it applied 8 newtons whether the target was a steel can or a strawberry. I implemented a closed-loop grip that ramped force until slip was detected by a sudden change in the sensor derivative, then held 15 percent above that point. Grasp success rose to 9 of 10, and the strawberry survived.3That project taught me the limit I now want to cross. Slip detection works for rigid grasps, but deformable objects change shape as you hold them, so the model of the object has to update in real time. This is exactly where the program's emphasis on cognition matters to me. I read Kappassov and colleagues on tactile sensing in manipulation and realized the open problem is not better sensors but better internal models, the kind biological systems build effortlessly.4My preparation maps onto the program's structure. On the mechatronics side, I built and tuned the gripper hardware and wrote the low-level firmware in C. On the machine learning side, my bachelor thesis trained a convolutional network to predict grasp points from depth images, reaching 78 percent success on the Cornell grasping dataset, and I learned the hard way why the remaining 22 percent failed on transparent objects. On the neuroscience side, I completed an online course in computational neuroscience and a seminar on predictive coding, which is why the chair of Prof. Cheng's work on neuroengineering drew me to TUM specifically.5I am aware of what I still lack. My mathematics is solid in linear algebra and probability but thinner in the optimization theory behind modern control, so I have been working through Boyd and Vandenberghe's Convex Optimization this spring, completing the first six chapters and the associated exercises. I mention this not to impress but because I want the committee to know I assess my own gaps honestly and close them without being asked.6After the master I intend to pursue a doctorate in robotic manipulation and then work in a research group, in industry or academia, on robots for clinical and care settings, where handling fragile and deformable objects gently is not a convenience but a safety requirement. TUM's combination of the Munich Institute of Robotics and Machine Intelligence, the Geriatronics initiative in Garmisch, and a faculty that treats cognition and engineering as one problem is the most direct path I have found to that goal.7I would be glad to contribute the manipulation experience and the habit of measuring failure precisely that I have developed, and to learn from researchers who have spent careers on the questions I am only beginning to formulate. Thank you for considering my application.
Yours sincerely,
[Name]8
- 1Opens by naming the exact program and quoting its own framing (the three disciplines). This signals fit with one specific program immediately, which is the first thing TUM rewards, rather than generic enthusiasm for 'robotics'.
- 2Replaces adjectives with numbers and a concrete failure (6 of 10 trials, two specific objects). TUM explicitly rewards evidence over adjectives; a documented failure is more credible than a claimed triumph.
- 3Shows the technical reasoning chain (diagnosis, mechanism, measured result). This is subject substance: the reader learns what the applicant actually understands about control, not just that they are 'passionate' about robots.
- 4Connects personal experience to a named research frontier and a citation, then ties it back to the 'cognition' pillar of the program. This three-way link (my work, the literature, the program structure) is what 'fit' looks like when it is earned rather than asserted.
- 5Walks through abilities discipline by discipline, each backed by a concrete artifact or metric, and lands on a specific TUM chair. This answers the prompt's four-part demand (abilities, talents, interests, ambitions) with subject substance rather than a personality portrait.
- 6Admitting a specific weakness and showing concrete remediation builds credibility. It also demonstrates self-directed rigor, signaling the applicant can survive a demanding research master without being managed.
- 7States a specific, plausible long-term ambition and ties it to named TUM institutes, closing the fit argument. The ambition is concrete (clinical and care robots, deformable handling) and follows logically from the gripper story, so the whole letter reads as one continuous line of reasoning.
- 8Closes briefly and in the applicant's own steady voice, offering something concrete (manipulation experience, rigorous measurement) rather than thanking effusively. The restraint matches TUM's preference for substance over self-promotion.
Stuck? Start here
- Which two or three first-year modules of this exact program connect to something I have already built, studied, or worked on?
- What is the single project or experience I could talk about for ten minutes under questioning without running out of detail?
- Why TUM specifically, and not just 'a strong German university', in one honest sentence?
Before you submit
- The letter is at most two A4 pages and, for a German-taught program, written in correct German.
- Every talent or strength is backed by a concrete example with a result or an insight.
- I have named real features of the program and answered why this degree and why TUM, and I can defend every claim in an interview.
Drafted it? Get an honest, admissions-style read, free.
Score my essay