西门子西门子数字化工业集团 物理AI平台 AI算法开发

工作描述

About the role
As an AI Algorithm Developer, you will help develop, validate, and productize intelligent algorithms for our Physical AI stack in China. You will work with the existing Suzhou Physical AI team on Level 2 robotics skills, including picking and placing, packing, palletizing, motion planning, perception, and adaptive robot behavior.
This role combines machine learning research and engineering with reinforcement learning and robotics application development. You will turn algorithmic ideas into measurable, maintainable capabilities that can be integrated, tested, deployed, and improved in real industrial environments.
If you enjoy solving complex perception, decision-making, and control problems with Python, machine learning, and reinforcement learning, this role offers the opportunity to shape practical Physical AI solutions.

Key responsibilities
•	Designing, implementing, training, and evaluating machine learning and reinforcement learning algorithms for robotics and Physical AI use cases
•	Developing perception, decision-making, planning, policy-learning, and adaptive behavior capabilities for industrial robotics applications
•	Building reproducible data, training, simulation, evaluation, and experiment-management pipelines
•	Defining baselines, metrics, benchmarks, ablation studies, and acceptance criteria to measure algorithm performance and robustness
•	Working with simulation and real robotic systems to address sim-to-real transfer, domain variation, safety constraints, and operational reliability
•	Optimizing models and inference pipelines for latency, compute, memory, and deployment constraints in edge or industrial environments
•	Packaging algorithms as maintainable software components or services with clear interfaces, documentation, tests, and versioning
•	Analyzing failures, data quality, model behavior, and edge cases, then translating findings into systematic improvements
•	Collabora…