Sim-to-Real / Control Engineer

The Mission

At Iacon Robotics, we believe the reality gap is merely a mathematical optimization problem. As a Sim-to-Real Engineer, your objective is to translate billions of hyper-parallel physical state-transitions generated inside our LBM into sub-millisecond, deterministic control loops deployed on unyielding metal.

You will own the bridge between the digital environment and edge silicon. If you do not care about cache-misses, Jacobian sparsity, or 1000Hz jitter, this node is not for you.

Core Responsibilities

• System Identification: Develop and deploy highly parallelized algorithms to accurately identify unmodeled physical mass matrices, friction coefficients, and actuator decay over time.
• Differentiable Simulation: Port custom rigid body dynamics code into CUDA to maintain fully differentiable physical constraints during policy training.
• Control Loop Architecture: Build deterministic C++ actuation pipelines that unroll policy inference and inverse kinematics into deterministic execution blocks with <50µs jitter bounds.
• Adversarial Stress Testing: Engineer noise-injection layers (Domain Randomization) into the physics pipeline to guarantee zero-shot transfer for the Model-01 Humanoid.

Qualifications

• Ph.D. or deep industry equivalent in Robotics, Control Systems, or Computer Science.
• Absolute mastery of C++17/20 and modern low-level systems programming.
• Extensive experience working under the hood of engines like MuJoCo, Isaac Gym, or Bullet.
• Demonstrated success deploying algorithms on edge hardware (Jetson, TPUs, raw ARM).
• Mathematical fluency in Lagrangian mechanics, operational space control, and whole-body MPC.

The Iacon Standard

We offer competitive baseline compensation and hyper-aggressive equity structures. We expect output that exceeds the boundaries of traditional software engineering. Deploy, or step aside.