About this role
The Cognitive and Human-like AI group at A*STAR’s Institute of High Performance Computing in Singapore seeks a scientist to research embodied artificial intelligence for robots that collaborate with humans on real-world tasks. The team develops collaborative intelligence for applications in manufacturing, trade and connectivity, with an emphasis on safety, commonsense reasoning and continual learning in complex human environments.
Responsibilities
- Research human-robot interaction, including multimodal perception, understanding of people and situations, world representation, decision-making, learning and behaviour.
- Design embodied AI algorithms for capabilities such as human-centric spatial awareness and intent prediction, supporting safe and proactive collaboration in shared workspaces.
- Develop, experiment with, test and integrate human-robot interaction modules for real-world robotics applications.
- Collaborate with researchers on scientific writing, algorithm development, testing and deployment of intelligence systems.
Required qualifications
- PhD in Robotics or a field related to embodied AI, such as Computer Vision, Natural Language Processing, Machine Learning, Cognitive Science, Cognitive Architectures or Robotic Foundation Models.
- Good programming skills in Python and/or C++, along with communication skills, teamwork and resourcefulness.
Advantages:
- Experience with ROS, Linux and/or other robotic middleware.
- Experience curating datasets and fine-tuning Vision-Language-Action or Vision-Language Models for robotics.
- Past work in 3D vision and spatial reasoning; human-robot interaction; representation of shared human-agent tasks or spaces; or modelling and reasoning about human actions and behaviours.
Skills for this role
Embodied Artificial IntelligenceHuman-robot interactionMultimodal perceptionComputer visionNatural language processingMachine learningSpatial reasoningIntent predictionRobot Operating System (ROS)LinuxPythonC++Dataset curationVision-Language-Action modelsVision-Language ModelsModel fine-tuningScientific writingSystem integrationCommunicationTeamwork