1 min read
Add as a preferred source on Google

Industrial Next Says Robot Cells Learn Tasks Through Simulation

The company advertises first-pass completion of 99.9% or higher and says one production task has a cycle time below 3.3 seconds.

A robotic gripper guides a cable into a smartphone chassis / TokenPost.ai
A robotic gripper guides a cable into a smartphone chassis / TokenPost.ai

Industrial Next says manufacturers can teach its robotic work cells new tasks through simulation and demonstration, an approach the startup aims to use to reconfigure production lines in minutes.

The San Francisco company describes its technology as models that connect vision, task context and robot motion. It says new tasks can be learned in hours and advertises first-pass completion of 99.9% or higher.

Industrial Next says one production task has a cycle time below 3.3 seconds.

The company’s technology is aimed at delicate manufacturing work. Its cells have been used for smartphone assembly tasks including threading a flex cable and placing a small connector cap. The phone customer has not been identified.

Co-founders Lukas Pankau and Allen Pan started Industrial Next in November 2021. Pankau said simulation has helped the company advance on precise, dexterous tasks. He described a longer-term goal of building a full car end-to-end without human involvement.

Simon Yoon

Reporter

Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to info@tokenpost.com.

Loading…