Role Summary

You will design the real-time sensing-to-action layer of Devyantram's robots, bridging perception, planning, control, and learning. This role owns how the CEREBRION bionic brain interacts with the physical world through sensors and actuators, and how the robot stays stable, precise, and adaptive in the loop, from single embodiments to coordinated multi-robot and swarm systems. A defining feature of this role: the cognitive layer is not a conventional digital controller. Its outputs come from biological (organoid-based) and engineered neural substrates coupled with optical processing, and a central part of your job is bridging that unconventional, non-deterministic compute to deterministic real-time control.

Key Responsibilities

• Architect end-to-end perception to control pipelines.
• Lead sensor fusion across vision, depth, tactile, and proprioceptive modalities.
• Develop closed-loop control for manipulation, locomotion, and balance.
• Apply both learning-based and classical control methods (MPC, whole-body control, RL hybrids).
• Own the ROS2-based control stack and simulation-to-real transfer.
• Integrate outputs from the platform's cognitive compute layer, which combines biological (organoid-based) and engineered neural substrates with optical processing, into deterministic real-time control loops, and own the rate, latency, and uncertainty mismatch across that boundary.
• Design multi-robot and swarm coordination: distributed estimation and control, multi-agent planning, consensus, and formation or collective behaviors across many agents and scales.
• Drive performance benchmarking in simulation and on hardware.
• Generate core controls and autonomy IP in coordination with patent counsel.

Required Qualifications

• MS or PhD in Robotics, Controls, EE, CS, or a related field.
• 8+ years in robotics, controls, or real-time autonomy, including deployment on physical systems.
• Strong foundation in control theory, robot dynamics and kinematics, and real-time systems.
• Excellent Python and C++.
• Deep ROS2 experience.

Strongly Preferred

• Humanoid or legged-robot control.
• Whole-body control, MPC, or task-space inverse dynamics.
• Vision-based manipulation and navigation.
• Experience with Isaac Sim, MuJoCo, or Gazebo.
• Hybrid classical-plus-learning control systems.
• Unconventional control systems, including bionic approaches.
• Multi-agent or swarm robotics, distributed control, or multi-robot coordination.
• Decentralized estimation and control (consensus, distributed MPC, or similar).

A Successful Candidate Will Be Measured By

• Stability, precision, and adaptability of robot behavior.
• Latency and robustness of closed-loop performance.
• Coordination, robustness, and scalability of multi-agent and swarm behavior.
• Clean integration across the cognition-to-actuation boundary, including the bridge from the unconventional cognitive substrate to real-time control.

The pay offered for this position may vary based on several individual factors, including job-related knowledge, skills, and experience. The total compensation package may also include additional components/benefits depending on the specific role. This information will be shared if an employment offer is extended.