AI/ML Research Engineer - Robotics Perception

SGInnovate; Singapore; On-site; SGD 90,000 - 140,000; Full time

Job summary

Vilota, supported by SGInnovate, seeks curious engineers to advance visual localisation and VIO algorithms for robotics. You will design state estimation, sensor fusion, and nonlinear optimisation, and prototype in Python or C++ for deployment on embedded platforms.

Ideal candidates have a strong math background, ML/CV knowledge, and hands-on experience with ROS, OpenCV, and SLAM. This role offers exciting, production-focused challenges in a cutting-edge startup environment.

Qualifications

• Bachelor's or Master's degree in Math, Computer Science, Robotics, or a related field.
• Proficiency in Python, C++, or Rust with strong software fundamentals.
• Solid knowledge of SLAM, perception, and sensor fusion concepts.
• Experience with ROS, OpenCV, Eigen, and GTSAM is highly desirable.
• Familiarity with ML and CV algorithms and 3D perception concepts.

Responsibilities

• Research, develop, and improve visual localisation and VIO algorithms.
• Design and implement state estimation, sensor fusion, and nonlinear optimisation.
• Develop mathematical models for cameras, IMUs, and other sensors including calibration.
• Investigate localisation failures and improve robustness under real-world conditions.
• Apply machine learning and classical computer vision to boost perception performance.
• Design experiments, analyse datasets, and evaluate algorithm performance.
• Prototype new algorithms in Python or C++, optimize for production deployment.
• Collaborate with hardware and software engineers to integrate into embedded platforms.
• Stay current with CV, robotics, and ML research and assess applicability.

Skills

Python

C++

Rust

Machine learning

Computer vision

SLAM

Education

Bachelor's degree in Math/CS/Robotics

Master's degree in a related field

Tools

ROS

OpenCV

Eigen

GTSAM

Job description

This role is posted on behalf of Vilota, a startup supported by SGInnovate.

Background

Vilota develops spatial perception products for autonomous and tele-operated robotic systems, focusing on lightweight sensor hardware, advanced vision-based software, and reliable operation in indoor and outdoor environments. We bridge cutting-edge research with production-ready systems, transforming theoretical algorithms into reliable products deployed in real operating environments. We are looking for curious, mathematically strong engineers or scientists who enjoy solving difficult problems and translating ideas into robust implementations.

Job Scope

• Research, develop, and improve visual localisation and visual-inertial odometry (VIO) algorithms.
• Design and implement state estimation, sensor fusion, and nonlinear optimisation algorithms for robotics applications.
• Develop mathematical models for cameras, IMUs, and other perception sensors, including calibration and error modelling.
• Investigate challenging localisation failures and develop practical solutions for robustness under real-world conditions.
• Apply machine learning and classical computer vision techniques where appropriate to improve perception performance.
• Design experiments, analyse datasets, and evaluate algorithm performance using quantitative metrics.
• Prototype new algorithms in Python or C++, and optimise them for deployment in production systems.
• Collaborate with hardware and software engineers to integrate algorithms into embedded and robotic platforms.
• Stay current with the latest research in computer vision, robotics, state estimation, and machine learning, and evaluate their applicability to company products.

Required Qualifications

• Bachelor's or Master's degree in Math, Computer Science, Robotics, Electrical Engineering, or a related field.
• Proficiency in programming languages such as Python, C++, or Rust.
• Strong understanding of sensor technologies (e.g., cameras, IMUs) and SLAM techniques.
• Experience with perception frameworks and libraries (e.g., ROS, OpenCV, Eigen, GTSAM).
• Solid knowledge of machine learning and computer vision algorithms
• Familiarity with 3D point cloud processing and geometric reasoning

Desirable Qualifications

• Experience deploying perception systems on drones, autonomous robots, or embedded platforms.
• Experience with camera and IMU calibration.
• Familiarity with deep-learning frameworks such as PyTorch or TensorFlow.
• Exposure to software optimisation for resource-constrained or real-time environments.
• Experience reading, implementing, or reproducing algorithms from academic research papers.