Senior Robotics Localization Engineer
Posted:
17 November 2026 (4 hours ago)
Application Deadline:
14 February 2027
Vacancies:
1 Vacancy
Job Summary
Analog builds robots-as-a-service for governments and enterprises and our neural agent Ana will eventually get a robotic body of her own. Neither works if a robot gets lost. Were 80 people in one office in Abu Dhabi and our robots are deployed and tested on-site right where we work. Youd own the layer that tells them where they are and your work would show up on a real robot not in a backlog.
**What were looking for and why**
----
Someone who **has shipped SLAM on real robots** and knows the gap between a lab demo and a robot that still knows where it is on day 90. Our robots mostly work indoors today so thats where we need it rock solid first.
**What youll be working on**
* Building maps including multi-session maps so robots can come back to the same space again and again
* Fusing LiDAR cameras and IMU so position tracking stays accurate
* Global relocalisation: a robot wakes up anywhere in a known map with no prior pose and works out where it is
* Local relocalisation: catching drift or tracking loss and recovering against the map
* Down the road outdoor localisation with GNSS / RTK
**What makes this role different**
---
Youll work on real hardware with no layers between you and the people building the rest of the robot and youll shape how the localisation stack grows indoors first and then outdoors.
**Skills we need**
* 5 years in SLAM / state estimation for mobile robots deployed on real hardware
* LiDAR and visual-inertial odometry; LiDAR-visual-inertial SLAM
* Multi-sensor fusion (factor graphs EKF)
* Map building and multi-session maps
* Global relocalisation (start-up and recovery in a known map no prior pose)
* Local relocalisation (detecting drift or tracking loss and recovering)
* ROS 2 C Python
**Nice to have**
* Outdoor state estimation: GNSS / RTK with LiDAR and IMU fusion
* Legged or humanoid robots (odometry with body motion)
* Semantic maps and map tools
**What were not looking for**
-
* People whose SLAM work has only lived in simulation or papers
* People who want to manage rather than build
* People who want a lot of process or a remote setup. Were in person in Abu Dhabi
**What were looking for and why**
----
Someone who **has shipped SLAM on real robots** and knows the gap between a lab demo and a robot that still knows where it is on day 90. Our robots mostly work indoors today so thats where we need it rock solid first.
**What youll be working on**
* Building maps including multi-session maps so robots can come back to the same space again and again
* Fusing LiDAR cameras and IMU so position tracking stays accurate
* Global relocalisation: a robot wakes up anywhere in a known map with no prior pose and works out where it is
* Local relocalisation: catching drift or tracking loss and recovering against the map
* Down the road outdoor localisation with GNSS / RTK
**What makes this role different**
---
Youll work on real hardware with no layers between you and the people building the rest of the robot and youll shape how the localisation stack grows indoors first and then outdoors.
**Skills we need**
* 5 years in SLAM / state estimation for mobile robots deployed on real hardware
* LiDAR and visual-inertial odometry; LiDAR-visual-inertial SLAM
* Multi-sensor fusion (factor graphs EKF)
* Map building and multi-session maps
* Global relocalisation (start-up and recovery in a known map no prior pose)
* Local relocalisation (detecting drift or tracking loss and recovering)
* ROS 2 C Python
**Nice to have**
* Outdoor state estimation: GNSS / RTK with LiDAR and IMU fusion
* Legged or humanoid robots (odometry with body motion)
* Semantic maps and map tools
**What were not looking for**
-
* People whose SLAM work has only lived in simulation or papers
* People who want to manage rather than build
* People who want a lot of process or a remote setup. Were in person in Abu Dhabi