Robotics Research Engineer

Maxwell Bond; Los Angeles (CA); Presencial; USD 90 000 - 130 000; Tempo integral

Resumo da oferta

Maxwell Bond is seeking a researcher to advance robotic manipulation through learning-based approaches. You will tackle challenging problems across data collection, model training, evaluation on real robotic platforms, and deployment of reliable systems.

You will work on one or two-armed manipulation, developing policies, curating datasets via teleoperation or simulation, and translating experimental results into practical improvements for robust robotic capabilities.

Qualificações

• Strong research experience in robotics, ML, or related field.
• Experience with robotic manipulation, imitation learning, reinforcement learning.
• Experience collecting and processing robotics data and evaluating on real or simulated systems.
• Applied research mindset translating results into practical improvements.

Responsabilidades

• Develop methods to collect, process and curate robotic training data via teleoperation and simulation.
• Design and train robot policies for manipulation tasks.
• Evaluate models on physical robotic systems and improve their performance in real-world conditions.
• Analyse deployed systems to identify limitations and implement improvements.

Conhecimentos

Robotics

Machine learning

Research

Imitation learning

Formação académica

PhD advantageous

Descrição da oferta de emprego

Help advance the next generation of robotic manipulation, developing intelligent systems that enable robots to carry out complex physical tasks. You'll work on challenging research problems and contribute to turning promising approaches into reliable performance on real robotic platforms.

What's in it for you

• Work on cutting-edge robotics research with direct applications to physical systems.

• Contribute across the development lifecycle, from data collection and model training to real-world deployment.

• Tackle challenging manipulation problems involving different robotic configurations.

• See your research translate into measurable improvements in robot capability and reliability.

The role

You'll research and develop learning-based approaches that enable robots to perform manipulation tasks using one or two arms. Your work may span the full pipeline, from acquiring training data and developing models to evaluating performance and improving deployed systems.

Responsibilities may include:

• Developing methods to collect, process and curate robotic training data using approaches such as teleoperation and simulation.

• Designing and training robot policies that perform manipulation tasks reliably and consistently.

• Evaluating models on physical robotic systems and improving their success rates in real-world conditions.

• Analysing the performance of deployed systems, identifying limitations and implementing improvements.

Experience needed

• Strong research experience in robotics, machine learning, computer science or a related technical field. A PhD is advantageous but not essential.

• Experience with robotic manipulation, imitation learning, reinforcement learning or other approaches to learning robot behaviours.

• Familiarity with collecting and processing robotics data, developing learning-based models and evaluating performance on physical systems or in simulation.

• A strong applied research mindset, with the ability to investigate complex problems and translate experimental results into practical improvements.