Research Scientist in Ai/ml for Dynamics and Control (hybrid)

RTX RTX · Aerospace · east hartford, CT +1 · Engineering

Research scientist specializing in learning for dynamics and control for aerospace and defense applications. Focuses on designing novel control solutions using machine learning, including reinforcement learning and neural networks, with an emphasis on safety-critical systems and verification.

What you'd actually do

  1. Design and develop novel control solutions for aerospace and defense applications including, but not limited to, jet engines, missiles, autonomous vehicles and systems, avionics, aircraft power systems and air management, hypersonic vehicles, advanced manufacturing, and space systems;
  2. Work in a multidisciplinary setting, bringing system-level perspective to new cutting-edge technologies from multiple fields (autonomy, power systems, cyber security, mechanical systems, aerodynamics, thermal management)
  3. Lead and support externally and internally sponsored programs, write external and internal research proposals;
  4. Disseminate research results through reports, conference proceedings, and peer-reviewed articles, and developing intellectual property.

Skills

Required

  • Ph.D. in Mathematics, Physics, Computer Science or Engineering
  • Control theory (multivariable control, estimation, LQR/LQG, Kalman filters, optimization-based control, Model Predictive Control)
  • Machine learning for control (Reinforcement Learning, Sim2Real transfer learning, Control Barrier Functions)
  • Verification & Validation of AI/ML control laws
  • Neural-network representations of controllers and estimators (Physics-Informed Neural Networks)
  • Control-oriented modeling of physical systems (first principles, data-driven)
  • MATLAB/Simulink
  • Python
  • Pytorch
  • TensorFlow

Nice to have

  • Master degree with 5 years of industrial experience
  • Control Barrier Functions (CBF)
  • Hardware-in-the-Loop validation
  • Real-time/embedded implementation of control laws
  • Speedgoat, dSPACE, LabView/NIDAQ
  • FPGA programming
  • C/C++ programming
  • Multi-agent collaborative autonomy
  • Decentralized mission planning and execution
  • High-fidelity simulations
  • Hardware platforms
  • PX4 or ArduPilot autopilots
  • Software-in-the-loop simulations
  • Robot Operating System (ROS, ROS2)
  • Gazebo simulation
  • Open-source planning and perception software
  • Commercial UAV and UGV platforms
  • Neural and symbolic AI approaches
  • Resilient contingency management
  • Human-robot teaming
  • Manufacturing and inspection operations
  • Assurance for autonomy
  • Safety and certification in aerospace
  • Gas turbine engine modeling and control
  • Guidance, Navigation, and Control (aircraft, spacecraft, or missiles)
  • Hypersonic propulsion
  • Electric or hybrid-electric propulsion for aircraft
  • Control co-design
  • Large Language Models
  • Agentic control

What the JD emphasized

  • U.S. citizenship is required
  • Ph.D. in Mathematics, Physics, Computer Science or Engineering
  • Strong fundamentals in control
  • Experience with machine learning for control
  • Reinforcement Learning (RL) for safety-critical systems
  • Verification & Validation of AI/ML control laws
  • Neural-network representations of controllers and estimators

Other signals

  • learning for dynamics and control
  • reinforcement learning for safety-critical systems
  • verification & validation of AI/ML control laws
  • neural-network representations of controllers and estimators