Perception Engineer III

Job description

At Zone 5 Technologies, we're redefining what's possible in unmanned aircraft systems. Our team of engineers and innovators is developing cutting-edge autonomous solutions that push the boundaries of UAS technology - solving complex challenges that matter. We're building the future of UAS capabilities, and we're looking for exceptional talent to join us. If you're driven by hard problems, energized by rapid innovation, and ready to make an impact on next-generation flight systems, you belong here. We are seeking a Senior Robotics Perception Engineer to serve as the perception technical lead embedded within a program team, developing seeker, target state estimation, and world modeling capabilities on top of our common AI/ML platform. You will own perception architecture and delivery for the program while contributing reusable capabilities back to the platform, working across the full pipeline from sensor data to actionable world state consumed by autonomy and mission software. **Responsibilities:** **Target State Estimation & Seeker Systems** - Design and develop target detection, tracking, and state estimation pipelines that produce robust track kinematics under challenging observation conditions - 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 - Balance 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) – equivalent industry experience also welcome - 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