Technical Program Head - Advanced Robotics Research

Job :
Berkeley, California, United States of America Salary: $203,000 - $348,000

Company:
Siemens
    Technical Program Head - Advanced Robotics Research
    
    Job
    
    523627
    
    Posted since
    
    23-Sep-2026
    
    Organization
    
    Foundational Technologies
    
    Field of work
    
    Research & Development
    
    Company
    
    Siemens Corporation
    
    Experience level
    
    Experienced Professional
    
    Job type
    
    Full-time
    
    Work mode
    
    Hybrid (Remote/Office)

Employment type

Permanent
    
    Location(s)
    
    Berkeley
- California
- United States of America

Reports to:

Global Head of Advanced Robotics Research.

Scope:
You own the robotics topic in the United States. You build and lead the California research team, set local technical direction, and coordinate with your peer Technical Program Heads in Munich and Shanghai under one global roadmap.

The Role Siemens is building a Long-Term Research organization with one mandate: create foundational technology that becomes Siemens business five or more years out. The robotics topic targets autonomous, poly functional robot systems: robots that generalize across tasks, embodiments, and industrial environments without per task reengineering.

Three sites carry distinct capabilities:
Munich, Shanghai, and California. You sit in the middle of the strongest robot learning ecosystem in the world, and your job is to make Siemens a serious research participant in it.

Siemens brings what the frontier labs do not have: real factories, real safety requirements, real deployment surfaces, and the automation stack that runs a large share of global manufacturing. Your models will be validated on physical industrial testbeds and run on hardware prototyped in Shanghai. Fragmented prototypes are a failure condition of this program.

You lead the California research program for Autonomous, Polyfunctional Robotics, with an initial focus on machine learning techniques for robots. You build and lead the team responsible for method development to enable robots to learn manipulation skills from human demonstrations using multi-modal robot foundation models, methods for the assurance of learned behavior, and the shared industrial benchmark, working with Munich and Shanghai under one global architecture and roadmap.

Research Agenda You Will Own Robot foundation models and policy learning. Research on Robotics Foundation Models including but not limited to vision language action architectures, world models, cross embodiment policy transfer, action space representations, and embodiment conditioned adaptation.

Learning from interaction for robotics. Create methods for robots to acquire and improve industrial skills through teleoperation, physical interaction, simulation, and autonomous experience. Establish data-generation and training approaches for coordinated use of arms, multi-fingered hands, torso, and locomotion, including manipulation while standing freely or moving under load.

Safety and assurance of learned whole-body behavior. Create the safety concepts, verification methods, runtime monitoring, simulation evidence, contact limits, fall and recovery envelopes, and certification arguments required for learned behavior in shared industrial spaces. The research object is not generic robot safety, but the assurance challenge created by learned whole-body motion, locomotion, balance, and physical interaction.

The robotics industrial benchmark. Lead the definition and evolution of a shared benchmark for industrial robotics work, including tasks, human baselines, acceptance metrics, trial protocols, and evidence requirements. Use the benchmark to measure whether a robot can perform work in different environments and to compare architectures, models, embodiments, and training approaches across the global program.

One shared robotics architecture. Co-own the global architecture connecting the robot’s brain, body, hands, control, sensing, simulation, data schema, and evaluation harness. Define the intelligence-side interfaces and the shared action representation so that learning results can transfer across regional platforms without creating disconnected site prototypes.

The Agenda Will Move New topics will need to be defined as the field and Siemens strategy evolve; some directions above will be redefined and others stopped. You bring an open mind, fast moving and dynamic execution, and the combination of depth and agility to pivot when the…