Mid-level Deep Learning Researcher

Работа на полставки

Ереван

08 Октябрь 2026

Условия контракта Гражданский договор

Категория Анализ данных

Подать заявку

Описание работы

ACSE is looking for a mid-level deep learning researcher to work on a project, “Adjusting PID Coefficients of UAV Control Using Reinforcement Learning,” within the scope of the Afeyan Family Foundation grant.

To learn more about the project please refer to the paper RLDroneSim: A simulation platform for reinforcement learning-based UAV experiments

Reports to: Lead data scientist, Principal Investigator

Обязанности

Responsibilities:

Collaborating on the development and refinement of RL-based PID tuning algorithms.
Set up, run, and monitor experiments using RLDroneSim, ArduPilot SITL, and Gazebo.
Run experiments based on configurations and protocols developed by the research team.
Assist with experiments involving wind disturbances, payload variations, sensor noise, and domain randomization.
Writing efficient, clean, and well-documented code for both simulation and hardware deployment.
Debugging, testing, and optimizing control algorithms in both virtual and real-world UAV environments.
Supporting data collection and analysis for performance evaluation and system improvement.
Contributing to technical documentation and preparing materials for presentations and publications.
Participating in team meetings, design discussions, and code reviews to ensure progress and alignment with research goals.

Требования

Must:

B.S. or M.S. in Computer Science, Data Science, Robotics, Electrical Engineering, Control Systems or related field. Students in their final year of studies can apply as well, given a strong background.
Strong understanding of the Reinforcement Learning Concepts or Control Systems
Strong proficiency in Python
Experience working with Git and collaborative development environments.
Experience with reinforcement learning frameworks (e.g., Stable-Baselines3, Ray RLlib, OpenAI Gym).
Ability to independently run experiments, debug models, and document findings clearly.

Preferred:

Understanding of PID controllers and control systems theory.
Familiarity with UAV simulation environments (e.g., Gazebo, PX4, SITL).

Дополнительная информация

What the Research Assistant Will Gain

Practical experience in deep reinforcement learning, UAV simulation, and adaptive control.
Hands-on experience with RLDroneSim, ArduPilot SITL, Gazebo, Gymnasium, and Stable-Baselines3.
Experience conducting reproducible machine-learning experiments.
Mentorship from experienced researchers and data scientists.
Experience preparing research figures, technical documentation, and publication materials.
An opportunity to contribute to research publications and presentations, depending on the level and quality of contribution.

Please submit your CV and Cover letter (with the names of three referees) via AUA career platfrom here.

AUA is an equal opportunity employer and is committed to an active non-discrimination program within the institution