E&T news weekly #97 – we choose our favourite engineering and technology news stories from the past week


Friday 3 June 2016

Jade Fell Jade Fell, assistant features editor
Asus unveils Wall-E inspired household helper robot Zenbo

Here he is, the robot we’ve all been waiting for! Asus have developed a new robot, adorably named Zenbo, designed to help around the house by carrying out tasks such as playing with children, and reminding the elderly when it’s time to take medication. I know there have been a few robots like this in the news recently, but this one is seriously exciting because it only costs £457; less than a new sofa! And this is something that will take away the horrible burden of keeping your own child entertained, which, let’s be honest, has got to be one of the most annoying aspects of parenthood. Who knows, maybe if a certain mother at Cincinnati zoo had a Zenbo to keep an eye on her son a certain gorilla wouldn’t be dead right now, but that’s just speculation. My main interest in this little robot, though, is not that it can look after children or the elderly, that doesn’t really help me out much right now, I’m more excited for what this has in store for the future of affordable household robotics. I am really hopeful that forthcoming updates might enable this little guy, or another equally affordable robot, to help out with more extensive household tasks, like hoovering, mopping, and entertaining guests when you can’t be bothered to put on trousers.

Tereza Pultarova Tereza Pultarova, news reporter
Heart-monitoring tablet named best tech innovation for Africa

One in two Africans over the age of 25 suffers from hypertension and nearly 20 million suffer from cardiovascular disease that could lead to heart failure, according to estimates. Yet the majority lives in areas with only limited access to healthcare. The best they can get is a poorly equipped rural hospital and so their problems frequently go unnoticed for many years. A solution to the problem might arise from this year’s Africa Prize run by the Royal Academy of Engineering. A low-cost tablet that allows any village doctor to take a cardiogram of a patient and send it over the Internet to a specialist has been named the winner of this year’s contest. Developed by Cameroonian engineer Arthur Zang , the device is already being used in Cameroon, Gabon, India and Nepal.

Asus unveils Wall-E inspired household helper robot Zenbo

Zenbo is likely the next ‘must-have’ of all technology geeks. Unveiled by Taiwanese computer maker Asus at Computex, an IT technology show in Tapei, Taiwan, the robot, somewhat resembling Disney’s Wall-E, is not only cute but appears to be actually quite useful. It responds to voice commands in spoken language, can find information online and read them out loud, remind people when it’s time to take medication or read stories to children. It can also serve as a control hub for smart home appliances and monitor the house when no one is around. Asus said that it will offer all that for about £457, which is a price of a regular PC.

Jack Loughran Jack Loughran, news reporter
Google project releases computer-composed piano song

As part of Google’s on-going research into machine learning, it has started a new project, called Magenta, which is looking into songs entirely generated by computers. A 90 second song consisting of a basic piano melody backed by a drum machine has been released to show the project’s current capabilities. It sounds… OK. It starts with a repeating refrain that deviates slowly over the course of the piece to give the impression that it is going somewhere without ever really stirring any serious emotions. To be fair to this first effort, the project is in its very early days and maybe the song did not adhere to this reporter’s normally more classic-rock oriented musical sensibilities. Attempts at procedurally generated music have been made in the past, such as Brian Eno’s ‘Bloom’ app which translates random presses on a touchscreen into an ambient soundscape. The output was passable but the main requirement of ambient music is to be acceptable background noise, hardly the most difficult genre to simulate. The Magenta project goes a step further by trying to create succinct musical pieces with hummable melodies. But many of the best songs (created by humans) come from inside the artist’s soul, an emotional outpouring of their deepest fears, their hopes, successes or memories of times gone by. None of these experiences can ever be replicated in a computer. The human elements of the most successful songs are what bring their composers global success by rousing relatable feelings in the brains of their listeners. But without these emotions, which are based on human experiences, it seems unlikely that a computer could ever provoke the same response.

dominic-lenton  Dominic Lenton, managing editor
57km Gotthard Base Tunnel opens under the Alps

This massive engineering project, which we looked at in detail as it approached completion in November last year, was hailed as a demonstration of both European fraternity and distinct national identity when it opened this week. The world’s longest and deepest railway tunnel will speed up journeys along the continent’s main line, connecting the ports of Rotterdam in the north to Genoa in the south. Swiss Federal Transport Office director Peter Fueglistaler said that it’s also a significant achievement for his country. Conquering the Alps, he claimed, represents to the Swiss what conquering the oceans was to the Dutch. There’s plenty of evidence for this; the Swiss public, despite opposition at times from the government and parliament, supported the gargantuan rail project in a series of binding referendums in the 1990s.

Katia Moskvitch Katia Moskvitch, technology features editor
Google project releases computer-composed piano tune

AI is coming – and actually, it’s already here. Recently, artificial intelligence has managed to create compelling human art. A group of researchers from Google’s Magenta team, part of the Silicon Valley giant’s Brain team, has used its open-source TensorFlow AI engine to create a 90 second piano melody written entirely via machine learning. The tune was composed without any human input and has a distinctive melody with repeating refrains. “Magenta has two goals,” says a blog post written by the team. “First, it’s a research project to advance the state of the art in machine intelligence for music and art generation.” Perhaps in the not so distant future, robots will replace not only cashiers and train drivers, but also composers.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *