Senior Localization Engineer / Research Engineer, Robotics AI (Visual Navigation)
Job Description
Get to Know Our Team
The Robotics Technology team is a core part of Grab's long-term vision to build urban embodied AI. We take full ownership of the product lifecycle — from perception and navigation research to real-world deployment on our autonomous delivery fleet across Southeast Asian cities. This is a fast-moving, multidisciplinary environment where robotics researchers, ML engineers, and hardware specialists collaborate to solve practical challenges at scale.
Based in Singapore, you will have the opportunity to work on frontier autonomy research, deploy solutions in complex real-world environments, and directly shape the future of last-mile logistics.
Get to Know the Role
As a Research Engineer on the Spatial Intelligence team, you will advance the core visual navigation and VLN capabilities of our autonomous platforms. Your research will focus on enabling robots to understand, reason about, and navigate complex urban environments using vision-based and multimodal learning approaches. You will bridge cutting-edge academic research with production-grade deployment, working closely with systems engineers to bring novel algorithms into our operating fleet.
You will report to the Head of Engineering and work onsite at Grab's One North office.
The Critical Tasks You Will Perform
1. Visual Navigation & VLN Research (50%)
• Design and implement novel deep learning models for visual navigation, Vision-Language Navigation (VLN), and visual place recognition in real-world outdoor environments
• Develop and evaluate multimodal perception pipelines combining RGB, depth, semantic segmentation, and language-conditioned navigation signals
• Research robust scene understanding techniques for dynamic urban environments: pedestrian-dense streets, repetitive urban facades, and GNSS-degraded corridors
• Investigate end-to-end learning approaches and modular architectures for navigation policy learning, with a focus on generalisation across unseen environments
• Publish and engage with the research community; contribute findings back to internal and open-source repositories
2. Algorithm Integration & Validation (30%)
• Implement and adapt state-of-the-art VLN and visual navigation algorithms (e.g., R2R, REVERIE, EmbodiedBERT derivatives) for deployment on real robot hardware
• Build performance evaluation frameworks and benchmarks tailored to outdoor, last-mile delivery scenarios
• Collaborate with Perception and Planning/Control teams to integrate visual navigation outputs into the broader autonomy stack
3. Research Iteration & Collaboration (20%)
• Drive closed-loop dataset collection and annotation pipelines to support continuous model improvement
• Partner with hardware and systems teams on sensor selection, calibration, and data quality for vision-based pipelines
• Present research findings to internal stakeholders and contribute to the team's research roadmap
Requirements
Function: Information Technology
Experience Level: Associate