Doctoral researcher, Project: 'Physical manipulation planning with differentiable closed-loop manipulation primitives'

(salary grade E13 TV-L, under the reserve that funds are granted, starting not later than from 1.10.2021 / for 3 years / closing date for applications 20.11.2020, Ref SCIoI-C3-39)

The Technische Universität Berlin invites applications for a position (PhD candidate) for the Cluster of Excellence “Science of Intelligence”.

What are the principles of intelligence, shared by all forms of intelligence, no matter whether artificial or biological, whether robot, computer program, human, or animal? And how can we apply these principles to create intelligent technology? Answering these questions – in an ethically responsible way – is the central scientific objective of the new Cluster of Excellence Science of Intelligence (, where researchers from a large number of analytic and synthetic disciplines – artificial intelligence, machine learning, control, robotics, computer vision, behavioral biology, psychology, educational science, neuroscience, and philosophy – join forces to create a multi-disciplinary research program across universities and research institutes in Berlin. Interdisciplinary research projects have been defined (, which combine analytic and synthetic research and which address key aspects of individual, social, and collective intelligence.

Working field

Doctoral project “Physical manipulation planning with differentiable closed-loop manipulation primitives”


The aim of this doctoral project is to demonstrate physical reasoning and manipulation planning on real-world robotic systems. Previous methods were successful in computing potential manipulation sequences and motion paths. However, their robust and reactive execution is essential to bring these capabilities to the real world. This project aims to close the gap between the existing planning methods and real-world execution. We investigate the idea of differentiable manipulation primitives and extend the existing planning frameworks to compute sequences of robust controllers rather than a single path. We also investigate perceptual anchoring of the controllers, as well as reactive re-initiation of parts of the solution sequence to cope with execution failures. The target is to demonstrate the capabilities on escape room scenarios. The doctoral student will work closely on the real-world system an develop control methods and tools to integrate with sequential physical manipulation planning.


Applicants must hold a Diploma/Master’s degree in computer science or a related discipline and should have proven skills/background in following topics:

  • Interest in interdisciplinary research in the context of the Center of Excellence “Science of Intelligence”
  • Robotics, control or AI planning
  • Strong programming skills (preferably in C++, or Python)
  • Strong analytical skills (linear algebra, calculus, AI theory)

Application procedure

Candidates should upload their application preferably via the portal in order to receive full consideration.

Applications should include: motivation letter, curriculum vitae, transcripts of records (for both BSc and MSc), copies of degree certificates (BSc, MSc), abstracts of Bachelor-, Master -thesis, list of publications and one selected manuscript (if applicable), two names of qualified persons who are willing to provide references, and any documents candidates feel may help us assess their competence.