AI Researcher Intern (Embodied Robotics) - Dec 2026 or Jan 2027 onwards (3/6 months)

Leveraging Sony’s world-leading sensor technology, Pentas Vision is exploring the integration of AI and advanced sensing in the field of embodied robotics. We are looking for a motivated AI Research Intern to support ongoing research projects in simulation and real-world robotics, helping us validate new sensors and algorithms in embodied AI tasks.

This position is ideal for students who are passionate about robotics, multimodal perception, and simulation, and who want to gain hands-on experience in cutting-edge embodied intelligence research.

What you will be doing:

• Research Exploration: Contribute to research on embodied AI algorithms, focusing on how multimodal sensing (vision, tactile, IMU, event-based, etc.) can enhance robotic intelligence.

• Sensor-Algorithm Integration: Collaborate with the team to incorporate sensor signals into learning frameworks, studying their effect on control, manipulation, and perception.

• Simulation Studies: Design and conduct experiments in simulation platforms (e.g., MuJoCo, Isaac Sim, PyBullet) to evaluate and benchmark embodied AI approaches.

• Sim-to-Real Transfer: Support investigations into transferring policies from simulation to physical robots, including analyzing sim2real gaps.

What you should have: 

• Currently pursuing or recently completed a Master’s or Ph.D. in Computer Science, Robotics, Electrical/Mechanical Engineering, or a related field, with focus on AI, robotics, or embodied intelligence.

• Solid understanding of deep learning architectures, with exposure to reinforcement learning, imitation learning, or embodied intelligence.

• Knowledge of multimodal perception and sensor fusion (e.g., vision + tactile, vision + IMU).

• Hands-on experience in integrating sensors or working with robotic hardware / ROS pipelines (e.g., data synchronization, calibration, basic debug of hardware–software stack).

• Experience with at least one simulation framework (e.g., MuJoCo, Isaac Sim, PyBullet).

• Proficiency in Python, with familiarity in PyTorch or TensorFlow.

• Strong analytical and problem-solving skills, with the ability to work independently and collaboratively.

• Strong communication skills.

Preferred Skills

• Publications in relevant conferences (e.g., ICRA, IROS, CoRL, RSS, CVPR/ICCV) or strong research project record in embodied AI / robotics.

• Participation in robotics competitions / AI challenges or demonstrable open-source contributions related to robotics/embodied AI.