3D Perception & SLAM Engineer

Weave Robotics•San Francisco, CA

About The Position

Weave is hiring a 3D Perception & SLAM Engineer to build the localization, mapping, and navigation systems that help its robots understand their position and changing surroundings in real customer homes. Unlike robots in controlled, pre-mapped environments, Weave’s robots must localize reliably amid changing lighting, moving people and objects, visually sparse spaces, and environments that evolve over time. The role spans state estimation, sensor fusion, calibration, and perception infrastructure. The engineer will integrate cameras, IMUs, odometry, and other onboard sensing into an accurate, real-time understanding of the robot and its environment. The work combines application and research: advancing established techniques and investigating how new end-to-end learned methods can be applied in the robotics stack.

Requirements

  • At least 3 years of experience building vision-based SLAM and sensor calibration for real-world robotic applications.
  • Deep understanding of state estimation, sensor fusion, probabilistic estimation, and techniques such as Kalman filtering, factor graphs, bundle adjustment, or nonlinear optimization.
  • Strong experience with multi-sensor calibration, time synchronization, and integrating cameras, IMUs, odometry, and/or LiDAR into robust perception pipelines operating under strict real-time constraints.
  • Experience writing production-quality software in C++ or Python.

Nice To Haves

  • Experience with ROS, embedded systems, or high-performance inference frameworks.
  • Experience building localization systems robust to dynamic scenes, changing environments, sensor degradation, or long-duration operation.
  • Experience building SLAM systems for mobile manipulators.
  • Experience evaluating or integrating learned components for visual odometry, feature extraction and matching, depth, or place recognition.
  • Experience with learned or end-to-end approaches to visual odometry, localization, mapping, depth estimation, feature extraction and matching, or sensor fusion.
  • Experience with state-of-the-art 3D reconstruction.
  • Familiarity with visual geometry transformers, Gaussian splatting, or VLAs.
  • Contributions to open-source SLAM systems or published work in the area.

Responsibilities

  • Build and optimize a production-grade indoor visual SLAM and localization system by fusing cameras, IMUs, wheel odometry, and other onboard sensors for robust, real-time robot localization.
  • Develop sensor infrastructure for calibration, synchronization, and data fusion to provide accurate, low-latency perception across the robot platform.
  • Architect localization and mapping, from core estimation algorithms through sensor integration, for reliable operation in dynamic indoor environments.
  • Explore and evaluate new localization and mapping approaches, combining geometric methods with learned techniques and bringing promising research into production on real robots.
  • Improve the robustness, computational efficiency, and power consumption of perception systems designed to operate continuously in customers’ homes.
  • Diagnose localization failures using logs, recorded sensor data, simulation, and deployed hardware, and turn real-world edge cases into systematic improvements.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service