Berkeley/NVIDIA Propose 'Agentic Robotics' to Reduce Data Dependency
Overview
Researchers from UC Berkeley and NVIDIA introduced 'Agentic Robotics,' using a 'Graph-as-Policy' framework to optimize robot control via simulation rather than massive real-world data. While partner Ambi Robotics pilots the method for parcel sorting, experts caution it remains suited for structured industrial tasks rather than complex home environments.
Generated from attributed reports · Updated 1 days ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · robots: 15000 other · Basis not reported
“ol program and test a new approach . “This is when I got goosebumps,” Goldberg said. Berkeley-based Ambi Robotics, which specializes in logistics and warehouse order fulfillment, applied the new methods earlier this year to analyze parcel sorting data and test updates to robot behavior. Tests showed an increased capacity of over 15,000 packages sorted annually per robot; the change is now deployed across 30% of the c”
Exact source · revision 2Source owner not reported
Artifact availability · code: reported pending
“s of agents can quickly generate, test and refine robot programs that allow human engineers to interpret how the programs work for fine-tuning and verification.” Goldberg, who was recently inducted into the National Academy of Engineering , is presenting the new approach and results from months of testing this week at the Conference on Robot Learning . He and his team are also releasing open-source code for other rob”
Exact source · revision 2Source owner not reported
Report timeline
Follow attributed reports and material updates.
- Robotics — Web news discoverySignalRecent advances in AI are creating a 'paradigm shift' for robotics
UC Berkeley roboticist Ken Goldberg and NVIDIA researchers propose 'agentic robotics,' a new discipline where teams of AI agents generate, test, and refine robot programs using simulation to overcome data scarcity. The approach utilizes a 'Graph-as-Policy' framework to segment tasks into skills and iteratively improve them. While the method claims to enable rapid adaptation without extensive data collection, the article notes that home automation applications like laundry folding are not yet imminent, though industrial throughput improvements have been demonstrated by partner Ambi Robotics.
- Chinese robotics — Research news discoverySignalNew 'Agentic Robotics' Approach Reduces Data Dependency in Robot Programming
Researchers from UC Berkeley and NVIDIA propose 'Agentic Robotics,' a method using an AI agent framework called 'Graph-as-Policy' to decompose tasks, simulate failures, and optimize control programs without relying on massive real-world data. While logistics company Ambi Robotics is piloting this approach for parcel sorting, experts caution that the technology currently suits structured industrial tasks rather than complex home environments, requiring further cross-industry validation before becoming a universal solution.
Related events
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).