Lead Controls & Dynamics Engineer

Pawaac Drones; Bengaluru; On-site; INR 1,800,000 - 3,600,000; Full time

Job summary

PAWAAC Drones is seeking a senior controls engineer to architect the PX4 flight stack, moving beyond parameter tuning to physics-first control for high-altitude gusts and challenging environments.

You will lead the Controls & Navigation vertical, implement advanced strategies (LQR, MPC, or H-infinity), perform system identification from flight logs, and validate gains on field flights with SITL/Gazebo before real hardware. You will be on the field deploying code and iterating gains in minutes.

Qualifications

• Deep understanding of flight dynamics and control theory.
• Experience replacing PID with modern methods.
• Experience with PX4/uORB and SITL tools.

Responsibilities

• Architect PX4 flight stack and implement custom control allocation.
• Design and validate controllers (LQR/MPC/H-infinity) for non-linear dynamics.
• Perform system identification from flight logs.
• Improve EKF for GPS-denied or noisy environments.
• Field validation with on-site deployments and rapid gains tuning.

Skills

Flight stack architecture

Advanced control design

System identification

State estimation

Field validation

Tools

PX4

uORB

Gazebo

AirSim

Job description

At PAWAAC Drones, we don’t just build drones—we engineer "eyes in the sky" for defense, rescue, and security forces. From Army to NDRF, from ports to police forces, our Quadrotor and DeltaQuad UAVs have flown in some of the toughest, most mission‑critical environments.

We’re a tight crew of IITians, DGCA‑certified pilots, and makers who believe "good enough" is never enough. If you’re looking for a place where you’ll get your hands greasy, your code tested in the real world, and your ideas airborne you’ve just found it.

The Mission: Why We Need You

Standard flight stacks (PX4/ArduPilot) are built for general‑purpose flying. They fail where we operate.

Our drones face high‑altitude gusts, maritime salt spray, and aggressive maneuvers that break standard cascade PID loops. We need to rip out the default control allocation and replace it with physics‑first, robust control algorithms.

You will not just be a user of the flight stack; you will be its architect. You will lead the Controls & Navigation vertical, owning the mathematical soul of the aircraft.

Core Responsibilities

• Flight Stack Architecture: Modify and rewrite core modules of the PX4 flight stack in C++. You must move beyond simple parameter tuning and implement custom control allocation matrices and mixing algorithms.
• Advanced Control Design: Replace standard PID loops with modern control strategies (e.g., LQR, MPC, or H-infinity) to handle non-linear dynamics and aggressive disturbances.
• System Identification: Conduct frequency domain analysis to characterize our custom airframes. You will extract linear state-space models from flight logs to validate your control laws.
• State Estimation: Improve the EKF (Extended Kalman Filter) implementation to handle sensor noise in GPS-denied or magnetically noisy environments (like ports or ships).
• Field Validation: You will not sit in a lab. You will be on the field, deploying code, analyzing flight logs (ULog), and iterating on gains in minutes, not days.

(You will be expected to surgically modify the specific 'Rate Controller' and 'Attitude Controller' blocks shown in standard architectures like this.)

Technical Requirements

• Language: Native fluency in C++ (C++11/14/17). You write efficient, realtime‑safe code without memory leaks.
• Flight Dynamics: Deep understanding of rigid body dynamics, quaternions, and 6-DOF equations of motion.
• Control Theory: Strong fundamentals in classical and modern control theory (Root Locus, Bode Plots, Lyapunov Stability, State‑Space).
• The "PX4 Internal" Knowledge: You know how the uORB messaging bus works, how the mc_att_control module functions, and how to write a custom driver from scratch.
• Simulation: Proficiency in SITL (Software In The Loop) with Gazebo or AirSim to validate crash‑inducing code before it touches real hardware.

The Pawaac Fit (Non‑Negotiable)

• First-Principles Thinker: You don't just copy‑paste code. You derive the math first.
• Autonomous Operator: You can identify a stability issue, read a research paper on "Active Disturbance Rejection Control," and implement a prototype by the next morning.
• High Agency: You treat the drone as your own. If a connector looks loose, you fix it. If the code looks messy, you refactor it.

Compensation: Competitive & Discussed based on capability.