This content is not available in your region

Exoskeleton gives super strength, new facial recognition software beats the human eye

Exoskeleton gives super strength, new facial recognition software beats the human eye
By Euronews
Text size Aa Aa

A cobot is a “collaborative robot” or exoskeleton which can help people run or walk while carrying heavy objects.

In France the ‘Hercules’ exoskeleton has been developed over the last five years by RB3D, a robotics company specialising in strength assistance.

This is just the latest version in an ongoing development project by the firm, which is based in Auxerre, around 200 kilometres south of Paris.

RB3D engineer and project manager Aurelie Riglet explained how easy it is to use: “You put this model on from the front, and in the future it’ll be just like putting trousers on, so getting ready will be very fast. Other new features include improved flexibility, you can even squat down in exoskeleton – and we’ve made it more compact for easier storage.”

The idea is that construction workers wearing the exoskeleton could, for example, carry a 60 kilo bag of cement but it would only feel like five kilos. The latest version is designed to fit a wide variety of body sizes, with quick links allowing the wearer to put it on in less than a minute.

However, Amelie Charnay, a reporter for the technology website pointed out super strength does not come cheap : “I think that the price is a weak spot, because currently it will cost 30,000 euros, so not everyone will be able to afford it, for most people it will be a luxury. For industry however, it is affordable. And another weak point is maintenance, constantly recharging the batteries, which is not always easy. It wouldn’t be possible to use it all day at full speed, and it’s fragile.”

Hercules can run for up to four hours in standard mode. According to the developers the exoskeleton could be on the road by 2016.

Be seeing you

At a university in Hong Kong, computer scientists have come up with facial recognition software which is even more accurate than the record they set back in April 2014.

Their new computer algorithm outperforms humans in recognising whether two photographed faces are the same person regardless of changes in lighting, make-up and camera angles.

Professor Tang Xiaoou, from the Engineering Faculty at the Chinese University of Hong Kong, said: “A couple of months ago we got a result that actually surpassed the human performance in recognising faces. This, although, is kind of a symbolic achievement, but researchers have been working on this for over 30 years so it’s quite exciting for us.”

Tang’s research started out purely for fun when he was sorting through photographs of his son from birth to 10 years old. But now big companies like Google, Facebook, IBM, and Microsoft are also interested in facial recognition technology.

Xiaogang Wang, an assistant professor in the university’s Electronic Engineering Department, explained: “We trying to develop a computation model to try to detect and check and analyse the behaviour of a crowd and predict potential crowd disasters. But the traditional way, video surveillance, they only focus on a small number of objects in a very simple environment. But now we target thousands of objects in very complex environments.”

Developers hope the software could help law enforcement and security agencies to identify individuals among a crowd of thousands.