Skip navigation
A Gripping Problem
Article

A Gripping Problem

Humans are masters of dexterity. But robots are catching up.

Nature, 2018

Read offline


Editorial Rating

8

Qualities

  • Innovative
  • Scientific
  • Applicable

Recommendation

Robots contribute precise, efficient work to many industries, and their use continues to expand, fueled by new technology. Today’s robots build and pack products with amazing efficiency, but their ability to grasp and manipulate objects remains limited. Researchers address these critical challenges through repetitive machine learning, advanced software and soft, human-like hand designs. Those curious about both the possibilities and limitations of today’s industrial robots will be engaged by this overview of new developments.

Summary

Scientists have developed impressive functionality for robots over the past 50 years, but grasping and dexterity remain critical challenges.

Although robots function successfully in many industries today, the real world is not an assembly line, where actions are continuously repeated. Today’s robots are learning to interact intuitively with a variety of objects. Engineers and industry leaders predict these new capabilities will have a profound impact on society and commerce.

Researchers use machine learning and improved software to address the gripping challenge...

About the Author

Richard Hodson writes for Nature Outlook supplements. He gained an MA in science journalism from City University London, and studied biomedical science at King’s College London.