Position Overview
We're seeking a UAV Control Expert who combines reinforcement learning with classical control methods to build flight control systems for interceptor drones. You'll develop RL-based control policies, tune PID controllers in the field, and bridge the gap between simulation and real-world flight. This is a hands-on role where your models fly on real drones.
Key Responsibilities
- Develop RL-based control policies for interceptor drone maneuvers
- Train RL agents for hold-in-place and target pursuit tasks
- Implement sim-to-real transfer pipelines with domain randomization
- Tune PID controllers for drone stability in the field
- Design hybrid control architectures combining RL with classical PID
- Iterate and improve control models based on real flight test data
- Set up and maintain simulation environments for RL training
Requirements & Qualifications
Essential:
- Deep experience or education in control engineering
- Hands-on experience with reinforcement learning for control tasks
- PID tuning experience, ideally on UAVs
- Proficiency in Python and PyTorch
- Understanding of drone dynamics and flight controllers
Preferred:
- Experience with sim-to-real transfer and domain randomization
- Familiarity with ArduPilot or Betaflight firmware
- PhD (or candidate) in control engineering, RL, or robotics
- Experience with simulation environments (Isaac Gym, Gymnasium, etc.)
- Comfortable working in the field with real drones
Technologies & Skills
Reinforcement Learning
PID Control
Sim-to-Real
PyTorch
ArduPilot
Betaflight
Python
Gymnasium
Isaac Gym
Work Environment
- Tiny, agile startup: your ideas and work have direct impact
- Lots of freedom to explore and propose solutions
- Challenging, meaningful tasks at the frontier of AI and robotics
- If you solve the task, your solution will be deployed on real drones
- See your work make a real-world difference
Flexible Applications Welcome
Not sure if you're a perfect fit? Apply anyway! We welcome candidates with diverse skill combinations. If you have expertise in multiple areas (e.g., RL + computer vision, control theory + embedded systems, or any other combination), we're particularly interested. We value multi-talented individuals who can contribute across different domains.
Ready to Join Our Mission?
If you're passionate about drone control, reinforcement learning, and making a real impact, we'd love to hear from you. Send us your resume and a brief introduction about your experience and why you'd like to join our team.
Apply Now
Contact: jobs@dremian.com