Sim2Real ML Engineer for Autonomous Construction Robots

Gravis Robotics; Zürich; Vor Ort; CHF 120.000 - 170.000; Vollzeit

Zusammenfassung

Gravis Robotics is seeking a talented AI/ML Engineer to join the Autonomy team in Zürich, focusing on bridging the sim2real gap for autonomous controllers in heavy machinery. You will develop ML/RL models, decide architectures, and quantify the data and sim-to-real transfer quality by building rigorous validation methods, working closely with simulation and autonomy teams on real robotic systems.

This role requires a degree in CS/Robotics/ML, strong Python and PyTorch, and a track record

Qualifikationen

• Degree in CS/ robotics/ ML or engineering.
• Strong Python and PyTorch, solid git skills.
• Experience modeling time-series or dynamical-system data from large datasets.
• Strong analytical skills to design experiments and draw defendable conclusions.

Aufgaben

• Build ML models to bridge the sim2real gap for autonomous controllers.
• Decide architectures (sequence models, state-space, etc.) and back with data.
• Characterize unmodeled effects and gap sources relevant to sim2real transfer.
• Assess data needs and distributions to cover deployment scenarios.
• Define performance metrics and validation for model fidelity and sim2real transfer.
• Collaborate with autonomy and simulation teams; influence controllers on the machine.

Kenntnisse

Python

Time-series modeling

Dynamical systems

Analytical thinking

Ausbildung

Degree in Computer Science, Robotics, ML or Engineering

Tools

PyTorch

Git

Jobbeschreibung

Gravis Robotics is seeking a talented AI/ML Engineer to join the Autonomy team in Zürich, focusing on bridging the sim2real gap for autonomous controllers in heavy machinery. You will develop ML/RL models, decide architectures, and quantify the data and sim-to-real transfer quality by building rigorous validation methods, working closely with simulation and autonomy teams on real robotic systems.

This role requires a degree in CS/Robotics/ML, strong Python and PyTorch, and a track record