Autonomous Machines - ARC

Infosys; Bengaluru; On-site; INR 1,400,000 - 2,400,000; Full time

Job summary

Infosys in Bengaluru seeks an advanced researcher/engineer to design and implement SLAM, scene understanding, and autonomous robot control. You will work with diverse robotic platforms, developing perception and planning pipelines and validating them in high-fidelity simulators.

The role emphasizes publication-ready research, real-world deployment experience, and collaboration with cross-functional teams to advance embodied AI in robotics.

Qualifications

• Master's degree in Robotics, Computer Science, AI, ML, Electrical Engineering or related field.
• PhD in robotics or related field preferred.
• Publications in Tier-1 conferences and Q1 journals, patents or significant project contributions in robotics.

Responsibilities

• Design and develop SLAM algorithms for robust indoor/outdoor localization and mapping.
• Research and implement Vision-Language Navigation (VLN) models for instruction-guided autonomous robots.
• Develop perception pipelines using CV, sensor fusion, and multimodal AI.
• Integrate and optimize robotic software on ROS/ROS 2 and simulation tools.
• Evaluate robotic systems with simulation and real-world benchmarks.

Skills

Python

PyTorch/TensorFlow

ROS/ROS 2

Computer Vision

Reinforcement Learning

Education

Master's in Robotics

PhD in robotics (desired)

Tools

Isaac Sim

Gazebo

Omniverse

AirSim

MuJoCo

Job description

Job Description:

Embodied AI and Autonomous Agents, Semantic Mapping and Scene Understanding, 3D Computer Vision and Point Cloud Processing, Grasp Planning and Manipulation, Multi-Robot Coordination and Swarm Robotics, Model Predictive Control (MPC), Autonomous Exploration and Frontier Planning, Navigation using Vision Foundation Models (CLIP, DINOv2, SigLIP), Digital Twins and Sim-to-Real Transfer, Safety-Critical Autonomous Systems, Edge AI deployment on NVIDIA Jetson platforms, Warehouse Automation, Industrial Robotics, and Field Robotics. Design and develop Simultaneous Localization and Mapping (SLAM) algorithms for robust localization and mapping in indoor and outdoor environments. Research, implement, and evaluate Vision-Language Navigation (VLN) models for instruction-guided autonomous navigation. Develop and deploy Vision-Language-Action (VLA) models enabling robots to perceive, reason, and execute complex manipulation tasks. Build autonomous navigation solutions using path planning, motion planning, obstacle avoidance, and trajectory optimization techniques. Work with robotic platforms including wheeled robots, UGVs, UAVs, mobile manipulators, and drone-based manipulation systems. Develop kinematic and dynamic models, including forward kinematics, inverse kinematics, motion control, and whole-body planning for robotic manipulators. Integrate and optimize robotic software using ROS/ROS 2, sensor fusion frameworks, and distributed robotic architectures. Create high-fidelity simulation environments using Isaac Sim, Gazebo, Omniverse, AirSim, MuJoCo, or similar simulators for development and validation. Implement perception pipelines leveraging computer vision, visual foundation models, multimodal AI, and sensor fusion across cameras, LiDARs, IMUs, GPS, and depth sensors. Utilize the NVIDIA Physical AI ecosystem, including Isaac Lab, Isaac Sim, Isaac ROS, Cosmos, Jetson, and Omniverse for robot learning and deployment. Develop AI-driven robotic solutions using Python, deep learning frameworks (PyTorch/TensorFlow), and reinforcement learning techniques. Evaluate robotic systems using simulation and real-world benchmarks, focusing on robustness, safety, scalability, and deployment readiness. Essential Qualification: Master's in Robotics, Computer Science, AI, Machine Learning, Electrical Engineering, or related field. Desired Qualification : PhD in robotics field Publications in Tier-1 conferences and Q1 journals, patents, or significant project contributions in robotics, autonomous systems, embodied AI, or computer vision. Experience deploying autonomous systems on real-world robotic platforms.