PEARL at TU Darmstadt

Robot learning for
intelligent assistance.

We develop robots that perceive, learn, and act through interaction. Our research connects mobile manipulation, dexterous and bimanual skills, and structured embodied intelligence.

Explore our research

Meet the team

TIAGo++ simultaneously opening a cupboard and placing an object inside, from the WHOLE-MoMa paper

WHOLE-MoMa: mobile manipulationCoordinating the base and arms on TIAGo++

MiTaS robot end effector with vision and tactile sensing for contact-rich manipulation

MiTaS: tactile learningContact-rich manipulation through touch

Two robot arms with dexterous hands jointly lifting a red box in the SYMDEX demonstration

SYMDEX: bimanual manipulationAmbidextrous dexterous skills

Move

Mobile manipulation

Learning to coordinate mobility and manipulation for everyday environments.

Explore the projects

Interact

Dexterity and contact

Learning manipulation skills through touch, demonstrations, and bimanual coordination.

Read our papers

Understand

Embodied intelligence

Structured representations and learning from interaction for perception, planning, and control.

Our research vision

Latest from PEARL

9 Nov 2026
PEARL at CoRL 2026: papers, demonstrations and workshops
Five accepted papers, one oral presentation, two invited workshop talks, MiTaS demonstrations, PolyUMI at the UMI Arena, and two postdoctoral opportunities. Meet…
Read more →
15 Jun 2026
Paper accepted @ RA-L
Semantic-Geometric Task Representations for Bimanual Manipulation from Human Demonstrations to Robot Action Planning Franziska Herbert, Vignesh Prasad, Han Liu, Dorothea Koert, Georgia…
Read more →
31 Mar 2026
New Paper accepted @ RA-L!
Self-Supervised Multisensory Pretraining for Contact-Rich Robot Reinforcement LearningRickmer Krohn, Vignesh Prasad, Gabriele Tiboni, Georgia Chalvatzaki TLDR: Multisensory pretraining enhances RL for contact-rich…
Read more →