Perception Engineer III

Responsibilities

• Target State Estimation & Seeker Systems Design and Development
• Implement track initiation, maintenance, association, and termination logic with well-characterized uncertainty
• Develop and tune nonlinear state estimators (EKF, UKF, particle filters) for target motion modeling
• Characterize seeker performance across engagement geometries, environmental conditions, and target signatures
• Define confidence and covariance outputs that downstream autonomy can reason about
• World State Modeling & Multi-Source Fusion
• Architect the world state representation that fuses self-state, target tracks, terrain, threats, and mission context into a coherent model
• Integrate outputs from VIO/SLAM, sensor fusion, and target tracking into a unified perception picture
• Manage temporal alignment, coordinate frames, and data association across heterogeneous sensor streams
• Define perception-to-autonomy interfaces with clear semantics for state, uncertainty, and freshness
• Autonomy Integration & System Architecture
• Own the perception architecture for the program and drive alignment with autonomy, guidance, and mission software teams
• Design DDS topic structures and message contracts for perception outputs within a WOSA-aligned architecture
• Build ROS2 nodes and libraries that expose perception capabilities through clean, versioned interfaces
• Contribute reusable perception components back to the common AI/ML platform for use across programs
• Balanced program-specific customization against platform-level standardization
• Embedded Performance & Deployment
• Develop and optimize perception code in C++ for deterministic performance on embedded compute
• Profile compute, memory, and I/O to meet real-time constraints on resource-limited platforms
• Make architectural decisions about what runs on GPU, CPU, or dedicated accelerators
• Support integration, hardware-in-the-loop testing, and field validation of perception subsystems
• Technical Leadership
• Write technical design documents, RFCs, and trade studies that drive perception architecture decisions
• Mentor junior perception engineers through code review, pairing, and technical direction
• Represent perception in program-level technical reviews and cross-team planning
• Translate program requirements into perception subsystem requirements with measurable acceptance criteria

  Qualifications

  • Bachelor's in Robotics, Aerospace Engineering, Computer Science, Electrical Engineering, Applied Mathematics, or related field (Master's preferred)
  • 6-9+ years of experience in robotics perception, state estimation, or sensor fusion
  • Strong proficiency in modern C++ for real-time embedded systems, with an emphasis on determinism, memory discipline, and performance
  • Hands-on experience with ROS2 and DDS-based middleware in production or program-delivered systems
  • Demonstrable experience developing and tuning nonlinear state estimators (EKF, UKF, or particle filters) for real applications
  • Working knowledge of at least one relevant area: target tracking, SLAM, VIO, or multi-sensor fusion – with the ability and interest to grow into the others
  • Experience deploying perception software on embedded compute platforms
  • Strong foundation in linear algebra, probability, coordinate frames, and estimation theory
  • Track record of owning a technical subsystem within a program or product from architecture through delivery
  • Ability to communicate technical trade-offs to both engineering and program stakeholders
  • Preferred: Experience with seeker systems, target tracking, or guidance-adjacent perception in a defense context
  • Hands-on work with WOSA, MOSA, or other open architecture standards
  • Deep DDS experience including QoS design, discovery, and performance tuning
  • NVIDIA Jetson deployment experience including profiling, memory management, and platform-specific optimization
  • Experience with SLAM/VIO frameworks (OpenVINS, VINS-Fusion, ORB-SLAM, GTSAM, Ceres)
  • Familiarity with PX4 or ArduPilot autopilot integration and MAVLink
  • Understanding of multi-hypothesis tracking, JPDA, or MHT approaches
  • Background integrating classical perception with learned components (detection, segmentation, feature learning)
  • GPU acceleration experience (CUDA, TensorRT) for perception workloads
  • Experience with hardware-in-the-loop test infrastructure and flight test data analysis
  • Familiarity with model-based development workflows (Simulink, ROS2 code generation)
  • Publishations, patents, or prior program contributions in tracking, estimation, or autonomous systems