TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

robotics

A Digital Twin Turned a Semantic Grasp Into a Physical One

GraspTwin treated foundation-model suggestions as seeds, then tested nearby poses in randomized physics.

Published Updated Story ID: mp-2026-09-28-010
Read the complete editionStory JSON

Summary

GraspTwin treated foundation-model suggestions as seeds, then tested nearby poses in randomized physics.

GraspTwin begins with one RGB-D observation, builds a digital twin and asks a foundation model for task-relevant grasp proposals such as using a mug handle for pouring. Rather than executing those proposals directly, it searches nearby poses with Bayesian optimization and evaluates batches in domain-randomized physics. The authors report zero-shot real-world transfer in minutes and task-oriented grasp success up to 33% above compared pipelines. The figure is specific to their objects, robot, simulator and baselines.

Why it matters

GraspTwin treated foundation-model suggestions as seeds, then tested nearby poses in randomized physics.

Limits and context

No additional limitation was separately recorded.

Key claims

  1. GraspTwin treated foundation-model suggestions as seeds, then tested nearby poses in randomized physics.

    Evidence: source-2026-09-28-010

Sources

  1. arXiv preprint 2609.30543arXiv · primary research

Corrections

No corrections have been recorded for this story.