{"id":160831,"date":"2021-01-22T20:44:30","date_gmt":"2021-01-22T17:44:30","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/new-robomorphic-computing-method-aims-to-speed-up-slow-witted-robots\/"},"modified":"2021-01-22T20:44:30","modified_gmt":"2021-01-22T17:44:30","slug":"new-robomorphic-computing-method-aims-to-speed-up-slow-witted-robots","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/new-robomorphic-computing-method-aims-to-speed-up-slow-witted-robots\/","title":{"rendered":"#New \u2018robomorphic\u2019 computing method aims to speed up slow-witted robots"},"content":{"rendered":"

“#New \u2018robomorphic\u2019 computing method aims to speed up slow-witted robots<\/strong>”
\n<\/p>\n

\n MIT researchers have developed a new way of speeding up a robot\u2019s reactions. They call it \u201crobomorphic computing.\u201d<\/p>\n

The method is designed to shrink the gap between a robot\u2019s \u201cmind\u201d and movements by creating customized chips that serve its specific computing needs.<\/p>\n

The system\u2019s inventors believe it could enhance a variety of robotics app<\/a>lications \u2014 including frontline medical care of COVID-19\u00a0patients.<\/p>\n

\u201cIt would be fantastic if we could have robots that could help reduce risk for patients and hospital workers,\u201d said<\/a> Dr Sabrina Neuman, an MIT CSAIL graduate who designed the method.<\/p>\n

[Read:\u00a0How this company leveraged AI to become the Netflix of Finland]<\/em><\/p>\n

Neuman says there are three main steps in a robot\u2019s operation: perception, mapping and localization,\u00a0and plotting a course of action.<\/p>\n

Robomorphic computing attempts to hasten this process by producing\u00a0specialized hardware that performs computing tasks more efficiently.<\/p>\n

The user first inputs the physical parameters of a robot, which the\u00a0system translates into mathematical matrices.<\/p>\n

It then creates a hardware architecture that\u2019s specialized to\u00a0run calculations on the matrices. This results in a chip design that\u2019s customized to the robot\u2019s computing needs.<\/p>\n

The researchers say a chip they programmed from the system\u2019s specifications performed eight times faster than a CPU, and 86 times faster than a GPU.<\/p>\n

They now hope to fabricate a custom motion-planning chip for every robot.<\/p>\n

Neuman plans to automate the entire\u00a0robomorphic computing process, so users can just drag and drop their robot\u2019s parameters to produce a hardware description.<\/span><\/p>\n

\u201cI think that\u2019s the thing that\u2019ll push it over the edge and make it really useful,\u201d she said.<\/span><\/p>\n

You can read the research paper abstract<\/a> at the website for the ASPLOS conference, where Neuman will present the work this April.<\/p>\n

\n Published January 22, 2021 \u2014 17:44 UTC\n <\/p>\n<\/p><\/div>\n