Senior Perception Hardware Engineer

JOB SUMMARY

As a Senior Perception Hardware Engineer focusing on Sensor Calibration, you will own the critical link between the physical world and our robot's autonomy stack. Apollo relies on a complex, multi-modal perception suite to navigate and manipulate its environment.

Your mission is to ensure these sensors act as a unified, mathematically perfect system. You will design the physical fixtures, write the calibration algorithms, and build the scalable factory processes that define the intrinsic, extrinsic, and temporal calibration of our humanoid robots. You will be the resident expert on optical physics, 3D geometry, and sensor precision.

ESSENTIAL DUTIES AND RESPONSIBILITIES or KEY ACCOUNTABILITIES

Calibration Architecture & Algorithms

• Sensor Modeling: Define and implement robust intrinsic models for various optical sensors (e.g., pinhole, fisheye, Brown-Conrady, Kannala-Brandt) and inertial sensors (bias, scale factors, random walk).
• Extrinsic Calibration: Develop highly accurate sensor-to-sensor and sensor to kinematic base spatial transformations.
• Temporal Synchronization: Work with electrical and embedded teams to validate hardware-level timing. Ensure sub-microsecond synchronization across all perception streams using PTP (Precision Time Protocol), PPS, and hardware triggering.

Hardware & Fixture Design

• Precision Fixturing: Design the physical calibration targets (e.g., CharuCo boards, custom 3D geometries) and the lighting environments required for high-repeatability factory calibration.
• Mounting & Stability Validation: Collaborate with mechanical engineers to evaluate and optimize sensor mounting rigidity. Characterize calibration drift caused by thermal gradients, mechanical shock, and joint wear.

Factory & In-Field Scalability

• End-of-Line (EOL) Automation: Build the automated calibration pipeline for our manufacturing line, allowing non-engineers to reliably calibrate a multi-sensor humanoid head in minutes.
• Online/In-Field Calibration: Develop lightweight "auto-calibration" or self-diagnostic routines that allow Apollo to detect sensor drift in the field and dynamically adjust its extrinsic transforms without human intervention.
• Data & Analytics: Establish a data-tracking architecture to monitor fleet-wide calibration health, identifying batch variances or specific hardware degradation over time.

SKILLS AND REQUIREMENTS

• Mathematical Foundation: Deep understanding of 3D geometry, linear algebra, coordinate frame transformations (quaternions, rotation matrices, SE(3)), and non-linear optimization (least-squares, bundle adjustment).
• Software Proficiency: Strong programming skills in Python and C++. Extensive experience with computer vision libraries (OpenCV, PCL) and optimization frameworks (e.g., Ceres Solver, g2o).
• Hardware Intuition: Hands-on experience working directly with optical components, LiDAR mechanisms, and MEMS IMUs. You understand the physical limitations that cause optical distortion, rolling shutter artifacts, and IMU drift.
• System-Level Thinking: Ability to translate abstract autonomy requirements (e.g., "we need 1cm grasping accuracy at 1 meter") into concrete hardware and calibration tolerances.

EDUCATION and/or EXPERIENCE

• Education: MS, or PhD in Robotics, Electrical Engineering, Optical Engineering, Computer Science, or a related technical field
• Experience: 5+ years of industry experience in perception hardware, computer vision, or robotics, with a primary focus on sensor calibration and validation.

PHYSICAL REQUIREMENTS

• Prolonged periods of sitting at a desk and working on a computer
• Must be able to lift 15 pounds at times
• Vision to read printed materials and a computer screen
• Hearing and speech to communicate