Senior Perception & Prediction Engineer, Robotics

Grab; Singapore; On-site; SGD 140,000 - 210,000; Full time

Benefits offered by this job

Term Life Insurance

Medical Insurance

GrabFlex benefits package

Parental leave

Birthday leave

LASA volunteering leave

Grabber Assistance Programme

FlexWork arrangements

Job summary

Grab is seeking a Senior Perception & Prediction Engineer in Singapore to architect and ship systems that convert multi-sensor data into a real-time, predictive world model. You will contribute across perception, prediction and planning, with emphasis on multi-object tracking, open-set perception, and end-to-end temporal models.

On-site at Grab, collaborating with cross-functional teams to scale robotics fleets.

Qualifications

• 3+ years hands-on experience shipping or operating perception and/or prediction systems for robotics, AV or ADAS.
• Rich experience across classical/geometric and learning-based approaches with interpretable safety fallbacks.
• Proven depth in multi-object tracking with data association and Kalman filtering.
• Hands-on experience with generic/open-set perception and occupancy representations.
• Depth in motion forecasting, occupancy flow or streaming BEV, with uncertainty calibration.
• Fundamentals in geometry and multi-sensor systems with calibration know-how.
• Proficiency in C++ and Python for production-quality code.
• Experience navigating reproducible experiments and log-driven root-cause analysis.
• Self-motivated with enthusiasm for autonomous driving and embodied AI.
• Familiarity with AI tools, model lifecycles and deployment considerations.

Responsibilities

• Develop and improve multi-object tracking and data association, including learning-based tracking.
• Build generic/open-set perception with obstacle detection and occupancy representations.
• Develop motion prediction for pedestrians, cyclists, vehicles and other agents.
• Explore end-to-end temporal perception-prediction approaches with modular baselines and deployable fallbacks.
• Investigate VLM/VLA and embodied foundation models for open-vocabulary perception.
• Strengthen multi-sensor and temporal integration across camera, LiDAR, radar and IMU.
• Define quality gates across tracking and prediction, drive down regressions with fleet data.

Skills

Robotics perception

Multi-object tracking

Open-set perception

Motion forecasting

Multi-sensor fusion

C++ & Python

ROS 2

OpenCV

PCL

NVIDIA Orin / TensorRT

Tools

OpenCV

PCL

ROS 2

NVIDIA Orin

TensorRT

Ceres

Job description

Grab is seeking a Senior Perception & Prediction Engineer in Singapore to architect and ship systems that convert multi-sensor data into a real-time, predictive world model. You will contribute across perception, prediction and planning, with emphasis on multi-object tracking, open-set perception, and end-to-end temporal models.

On-site at Grab, collaborating with cross-functional teams to scale robotics fleets.