a.Asil Kaan
Bozcuoğlu

Robotics & embodied AI / Berlin, DE

From understanding
to action.

I’m Asil Kaan Bozcuoğlu, a robotics researcher and engineer connecting language, perception, and action.

I work on vision-language-action models and the simulation systems that help robots learn, reason, and operate in the physical world. My latest project, Grounded VLA, explores evidence-grounded, explainable robot execution.

Explore my work Read the research
Dr.-Ing. · University of Bremen13+ years in robotics & AI
PERCEPTION → LANGUAGE → ACTION
Asil Kaan Bozcuoğlu wearing a navy blazer and light blue shirt

Asil Kaan
Bozcuoğlu PhD

Berlin, Germany

LEARNING IN SIMULATION.ACTING IN THE WORLD.
Ideas → systemsπ₀ / SmolVLAO3DENVIDIA OmniverseIsaacPyTorch / ROS

01 / Research & engineering

Intelligence, grounded
in the physical world.

From learned robot policies to the virtual environments and knowledge systems that support them.

01

Vision. Language.
Action.

Research and development with VLA models, including π₀ and SmolVLA, for robotic task planning and execution. Connecting multimodal models with robot behavior.

VLA modelsRobot learningEmbodied AI
02

Simulation.
Digital twins.

Building embodied AI workflows with O3DE for training, evaluation, and sim-to-real research. Technical leadership in photorealistic, physics-enabled digital twins with NVIDIA Omniverse and Isaac.

O3DEOmniverseIsaacSim-to-real
03

Knowledge.
Reasoning.

Bringing learned models and structured knowledge together. From KnowRob and cloud robotics to language-based planning, action ontologies, and retrieval grounded in the robot’s environment.

Neurosymbolic AIOWL / RDFPlanning

02 / Publications

A research trail.
From memory to action.

Google Scholar Download BibTeX

26 publication records · newest first

Includes journal articles, conference and workshop papers, preprints, and theses.
Updated with the author’s Google Scholar list · September 2026. Sources & scope

03 / About

Research depth.
Engineering perspective.

I’m a Berlin-based researcher and engineer with more than 13 years of experience across robotics, artificial intelligence, and simulation.

My work spans academic research and industrial R&D: learning from robot experience, building cognitive systems, leading digital-twin development, and exploring foundation models for robotics.

I earned my Dr.-Ing. in Computer Science at the University of Bremen, researching fast robot learning through prospection, episodic memories, and symbolic knowledge.

Download CV (PDF)
English · C2German · C1Turkish · Native

Recent research & development / Huawei

Robotics & Embodied AI

VLA models, language-based planning, and O3DE simulation workflows for embodied AI training and evaluation.

2021–2023 / Deutsche Bahn

Digital twins & autonomous systems

Technical Lead for photorealistic, physics-enabled simulation at Digitale Schiene Deutschland. Led six engineers building a railway digital twin with NVIDIA Omniverse and Isaac, alongside decision-making research for autonomous trains.

2012–2021 / University of Bremen

Robot learning & cognitive systems

Doctoral and postdoctoral research at the Institute for Artificial Intelligence. Cloud robotics, robot memory, affordance modeling, teaching, and research supervision.

Education

Dr.-Ing. Computer Science · Bremen · 2019

M.Sc. Computer Engineering · METU · 2012

B.Sc. Computer Engineering · Bilkent · 2010

2021

Best Technical Paper

PETRA · Learning Visuomotor Policies with Deep Movement Primitives

2018

Best-paper finalist × 2

ICRA · Cognitive Robotics & Service Robotics
The Exchange of Knowledge Using Cloud Robotics

In practice

From code to robots

Python · C++ · PyTorch · ROS
Docker · Linux · SWI-Prolog · OWL / RDF

04 / Let’s connect

Let’s build what
robots do next.

Interested in embodied AI, robot learning, or simulation?
I’d be glad to hear from you.

abozcuoglu@gmail.com

Based in Berlin.
Thinking beyond boundaries.