2023-06-21 12:00:07
More than 6,000 researchers and industrialists gathered in London at the end of May for the International Conference on Robotics and Automation (ICRA), the annual high mass of robotics, combining presentations of the latest research results – in fields as varied as locomotion or human/robot interaction – and demonstrations ranging from quadrupeds Spot et Go1 to the winged robot Nimble.
On the material side, the image of the robot with the general public is often anthropomorphic. In 1996, after years of research conducted in the greatest secrecy, Honda unveiled P2, who looks like an astronaut in a space suit, is 1.80 meters tall and weighs more than 200 kilos, but above all, walks with an ease and naturalness never seen before. More and more evolved humanoids will lead in 2013 to Atlasfrom Boston Dynamics, a prodigy of mechatronics and design, capable today of dancing, jumping, pirouetting, etc.
Read also: Article reserved for our subscribers Ocean One, a humanoid robot in the abyss
It is nevertheless clear that none of these incredible machines, despite their media and academic impact, has (for the moment) had a real impact on the industry, and this despite the ambitions recently displayed by Tesla for Optimus.
Mainly articulated arms
Today, as sixty years ago, the majority of industrial robots are, in fact, articulated arms installed on assembly lines and devoted to repetitive tasks such as painting or assembly. They are built by a handful of large manufacturers, mainly located in Southeast Asia and Northern Europe, and no longer, paradoxically, in the USA, pioneers in this field at the end of the 1950s. Specialized robots find, on the other hand, many new applications, ranging from vacuum cleaners Roomba to mobile platforms Kivadeployed by the tens of thousands in Amazon warehouses.
Read also: Article reserved for our subscribers “Robotization benefits the productivity of companies that adopt these technologies and their jobs – all their jobs”
In terms of intelligence, robotics faces three major difficulties. First, it must integrate many aspects of modern artificial intelligence, including navigation, planning and perception. Next, a robot is a very real machine whose physical interactions with the environment can lead to material damage, which makes reliability and safety two primary issues and makes the exploration process extremely delicate, one of the keys to reinforcement learning, for example. In this context, the arrival on the market of sophisticated simulation platforms can have a significant impact.
You have 32.7% of this article left to read. The following is for subscribers only.
#Intelligent #robots #revolutionizing #industry