Ranking Member Lofgren's Opening Statement at Robotics Hearing

WASHINGTON, D.C.- Today, the House Committee on Science, Space, and Technology is holding a Research and Technology Subcommittee hearing titled, Robots Made in America: Advancing U.S. Leadership in Manufacturing and Automation.(link is external)

Ranking Member Zoe Lofgren's (D-CA) opening statement as prepared for the record is below:

Thank you, Chairman Obernolte and Ranking Member Stevens, for holding today's hearing. I would also like to thank our expert panel of witnesses for your testimony today.

Significant progress in artificial intelligence has led to major innovation in robotics. Computer vision, natural language processing, and reinforcement learning have accelerated robotics development and allowed robots to be deployed in new environments and use cases. AI techniques have also accelerated the intelligence of fixed robots, enabling people to safely and easily control and work alongside them. My district is half Silicon Valley and half agriculture, so I have seen robotics take root in both of these sectors.

Unfortunately, the significant policy interest in the software has not been matched by the policies that are needed to advance the full robotics system. Recent administrations - Democratic and Republican - have prioritized AI development while underinvesting in hardware development. The National Robotics Initiative, which was started by the Obama Administration in 2011 to coordinate federal investment in robotics R&D, was phased out in 2022. The current Administration's AI Action plan included just one recommendation on robotics development as part of a longer list of next generation manufacturing technologies.

While individual science agencies continue to make investments in robotics today, there is no coherent strategy for U.S. leadership. In the meantime, China has races ahead to claim that leadership position.

My message today is simple: to reclaim our leadership in robotics innovation and manufacturing, the United States need to invest in both the "brains" and the "bodies" of these systems. We must also invest in solving the tricky problems at the intersection of the two.

Let me provide an example. We have a long history of developing and implementing occupational safety standards for the use of fixed robotics systems that work alongside humans in our factories. However, many of these industrial safety protections do not account for advanced robotics systems that use AI and have variable decision making. Further, many companies are starting to talk about putting robots in places where no safety standards currently exist-including in our homes and in elder care facilities.

We need to have a serious conversation as a nation about what safety means for robotics deployed in these settings before they cause physical harm. It's bad enough when a delivery robot shatters a bus stop shelter. Imagine the consequences of having a 200-pound robot in your elderly parent's living space without strong safety standards.

Another major challenge is data. One of the main reasons for AI's advancement in recent years is the broad availability of data to train AI systems. Computer vision was helped by access to images on the internet, and large language model development was improved by access to Wikipedia. However, until relatively recently, there were few equivalents for training physical robotics systems.

Some U.S. organizations are trying to address this problem. Individual U.S. companies are developing data sets for their robotics systems. Over 21 academic institutions, including my alma mater Stanford University, have partnered to build large data sets of specific robotics tasks. However, this effort is being undertaken without direct federal support.

Compare these relatively limited efforts to the national effort going on in the People's Republic of China to generate robotics data. The country has launched data sharing initiatives, data factories where humans train robots by preforming tasks, and efforts to standardize training datasets and pipelines for robotics.

America still can and must lead in the field of robotics. But to achieve that goal, we need a concerted, national effort to support innovation across the full robotics system. I look forward to hearing from each of our witnesses about the challenges and opportunities in robotics during today's hearing.

Thank you and I yield back.

###