Controls Engineer

EARTHBRAIN, the construction-technology company, is developing fully unmanned construction, and is hiring a Controls Engineer to build the control and model-estimation side of the autonomous system. The engineer works with both the autonomous-control and AI teams rather than belonging to only one.

The work has two parts. In system identification and modelling, the engineer designs and runs real-machine data collection on machines such as bulldozers, estimates and verifies state-space models from the data, optimises model parameters statistically and checks accuracy in physics simulators such as AGX Dynamics and NVIDIA Isaac Sim. In control development, the engineer implements and tunes autonomous control, including model predictive control, designs cost functions, builds state estimators such as Kalman filters, and leads experiments on real machines.

Applicants need an understanding of state-space models, optimal control design such as LQR and LQG and state estimation, C++ real-time control implementation, Python control tools, about three years of system identification, simulator-based verification and control experience on real robots, construction machinery or vehicles. Nonlinear MPC, GPU acceleration, reinforcement learning with MPC and ROS experience are preferred.