I am a robotics engineer at ASELSAN. My research combines
reinforcement learning with model-based whole-body control for legged
robots, and my engineering work focuses on control and autonomy
software for physical robots.
Publications
ICRA 2027 · Under review
Learning Structured Locomotion References for
Optimization-Based Whole-Body Control
Ö. T. Banuş, U. Saranlı, M. M. Ankaralı
A reinforcement-learning policy adapts the references of a
whole-body impulse controller. In simulation it reached
77.29% mean survival under velocity disturbances, compared
with 50.35% for the model-based reference controller.
Learning Hierarchical Task-Space Control Policies for
Whole-Body Locomotion
Ö. T. Banuş · Advisors: M. M. Ankaralı, A. E. Turgut
A learned policy generates task-space references for a
model-based whole-body controller. It tolerated larger
horizontal pushes than a WBC+MPC baseline, which remained
stronger against vertical ones.
Contact Consistent Disturbance Estimation for Quadruped Robots
H. C. Özden, E. Tanfener, H. E. Orhon, Ö. T. Banuş, U.
Saranlı, M. M. Ankaralı, A. E. Turgut
Estimates external body forces and ground reaction forces from
proprioception. Evaluated in simulation and on a Solo-12
robot, for which I ported our in-house locomotion controller.
Descriptions of employer projects are limited to what can be shared
publicly. A more detailed portfolio is available on request.
ASELSAN · 2022–2025
Blind locomotion on an in-house quadruped
One of five engineers who developed ASELSAN's in-house quadruped
from its initial concept; responsible for its software and
locomotion controller. The robot has a heavy metal body and
walks over highly uneven terrain, slopes, and stairs without
exteroceptive sensing. Work included slope-aware MPC, contact
estimation from joint torques, and actuator-aware whole-body
control.
Deployed on three versions of the in-house quadruped, Ghost
Vision 60, Unitree Go1, Unitree Go2, and Solo-12.
ASELSAN · 2025–present
Onboard software for a high-speed autonomous UAV
Designed the ROS 2 software architecture and mission state
machine, the interface to the flight-management unit, and the
simulation environment used to test the full stack before
flight. Work on navigation, GNSS-denied state estimation, and
trajectory planning.
ASELSAN · System owner
Leader–follower quadruped on Ghost Vision 60
A quadruped that follows a person on foot, using UWB and IMU
fusion to estimate the leader's position and a LiDAR map with
Nav2 to follow. Wrote the system requirements, coordinated the
departments involved, and led integration through field
delivery. Demonstrated publicly at IDEF 2025.
Bilkent University · 2021–2023
Modelling a collision-resilient micro quadrotor
A hybrid flight and ground-contact model of a foldable
quadrotor with compliant landing gear, a RotorS simulator
extended with cascaded PID control, and a Gazebo plugin for the
compliance forces.
Specialist Robotics Systems and Control Engineer, ASELSAN
Oct 2025 – present
Onboard software for a high-speed autonomous UAV.
Robotic Systems and Control Engineer, ASELSAN
Dec 2022 – Oct 2025
Software and locomotion control for an in-house quadruped;
system owner of a leader–follower quadruped.
Undergraduate Research Assistant, Bilkent Miniature Robotics
Laboratory Oct 2021 – Jun 2023
Modelling and simulation of collision-resilient quadrotors under
Assoc. Prof. Dr. Onur Özcan.
Flight Controls Engineering Intern, Turkish Aerospace
Jun – Jul 2022
System identification with frequency-sweep inputs.
Robotics Intern, OPLOG / Tarqan Robotics
Feb – Jun 2022
Control, sensor fusion, and state estimation for mobile robots.
Teaching Assistant, Bilkent University Physics Department
Sep 2019 – Jun 2021
Laboratory instruction and grading.
Education
MSc in Robotics, Middle East Technical University
Sep 2023 – Sep 2026
Thesis advised by Assoc. Prof. Dr. M. Mert Ankaralı and Assoc.
Prof. Dr. Ali Emre Turgut, in collaboration with Prof. Dr. Uluç
Saranlı, at the Center for Robotics and Artificial Intelligence
(ROMER).
BSc in Mechanical Engineering, Bilkent University
Sep 2017 – Jun 2023
Full scholarship with monthly stipend; honor certificates for
academic excellence.
Skills
C/C++, Python, real-time Linux, ROS 2, Docker, MATLAB/Simulink,
MuJoCo, Gazebo, EnvPool, Cheetah Software. Model predictive control,
whole-body impulse control, optimal control, spatial dynamics,
contact modelling, state estimation, reinforcement learning (PPO),
hardware integration over CAN and RS-232.
Awards and activities
Team member, ASELSAN–METU/ROMER legged robotics collaboration,
which received the 2025 YÖK University–Industry Collaboration
Award.
Ranked 157th among approximately 2.5 million candidates in the
2017 national university entrance exam.
10th International Summer School on Screw-Theory-Based Methods in
Robotics, 2024.
Volunteer building supervisor, Oy ve Ötesi, 2018–2023.
Languages
Turkish (native), English (full professional proficiency), German
(elementary).
