Author Archives: nico hemp

Industry 4.0 – Adidas joins in

Reading Time: 2 minutes

Rumours have it that Adidas committed to move some production back to Germany as a by-product of the negotiations with DFB (German football association) sponsoring. Particularly the jersey of the national team was said to be produced in Germany again, 20 years after outsourcing that whole part of the value chain to Asia. That was about half a year ago, and ignorami like me seriously wondered how that would be possible. Well, like this:

If everything goes according to plan, Adidas can start the series production next year. A second factory is then built in America, the company management announces. Both sites together would produce one million pairs of shoes annually. A small amount – measured by the total volume of around 300 million pairs a year. But it should be the beginning. “In the medium term, we will find such factories in all major sales markets,” announced CEO Herbert Hainer. And he doesn’t conceal the motivation – it’s the logo ‘made in Germany/USA/Australia’ etc. that still signals quality apparently.

Shoes in the robot factory can be manufactured even more cost-effectively than man-made running shoes. Only the staff will differ: While traditionally sewed and patterned shoes in Asian factories require human intervention, the pilot factories in Ansbach and America each require not more than 160 production and maintenance specialists for the robots.

Adidas’ foray fits neatly in the returning trend towards automation. Industry 4.0 penetrates all sectors, also other industries profit. At the Audi factory in Ingolstadt, for example, people and robots are already working side-by-side: in final assembly, a robot gets the necessary components and passes them to the employee.

Then again, we’ve been there before. And actual research shows the limits of automation. Not necessarily in terms of feasibility but efficiency. If we think about the hundreds of millions Google, Tesla, BMW etc. have invested in autonomous driving (which essentially is robotics and still doesn’t work perfectly) and that Adidas in this example even received funding from the German government for its robot factory, the question pops up how long it will take until someone will invest that kind of money for a robot to cut meat (which from an engineering point of view is everything but simple) or to do housework.

The last shall be first

Reading Time: 2 minutes

In the shade of the company’s most recent innovations in the field of disguise, it is easy to overlook Volkswagen’s attempts to bring autonomously driving cars to the streets. Especially since everyone else seems to do the same thing and everyone else seems to do it so much better. Europeans love Tesla, Americans can’t get over the Mercedes F015 and “Google’s self-driving car” always delivers when a journalist doesn’t know what to write about.

Volkswagen is a bit late to the party. Last year they presented the Budd-E, an electronic van, that could drive autonomously if it existed, because what VW actually presented was a mere concept study. At a time when Google’s vehicles already got crashed by other cars, the Mercedes drove around in San Francisco and Tesla (S Model) and BMW (5er) have parts of the technology in their serial cars already.

One thing that is easily forgotten though is that autonomous driving will work infinitely better for every other car that also drives autonomously. The benefit for a city with only self-driving cars will be greater than the sum of benefits for the individual car owners, think traffic jams, resource efficiency, parking spaces and so on. And in case the future doesn’t hold Teslas, BMWs and Mercedeses for everyone, it will be Volkswagen’s job to deliver. Volkswagen analogously means ‘car for the people’. It is not so much about innovating but making innovation available and affordable. It is debatable whether its new 150 kW charger system is a step in that direction. In Germany it takes a while to find a 50 kW charger already, but it seems to be consensus among car producers to aim for 30 minute charging time for 80 percent battery life. There seems to be trust in the state concerning infrastructure development (within Europe that is).

The car Volkswagen presented at this year’s CES was basically a VW Bulli. A continuation of its old transporter series. My parents drove a fossilized, rusty version of that car when they were my age, the thought of the next generation driving a then old fully automated VW T11 on their road trips is actually more futuristic to me than a Swiss dentist in his Mercedes F. Innovation is for everyone.

NB: No camouflaged advertising for VW intended, everybody knows Toyota HiAce is the better car.

Making water

Reading Time: < 1 minute

Clean drinking water is scarce in many parts of the world – especially in hot areas. According to the WHO 780 million people worldwide do not have access to clean water, with 3.4 million deaths related to it.  Though in many dry regions in terms of precipitation, humidity is often particularly high. For these regions, the Israeli start-up Water-Gen has developed a device that draws water from the air. A power connection is required but the device is arguably more energy efficient than previous models because it uses already cooled air to reduce the temperature of incoming air. Thus far the company has worked with the military of various countries, but now the civilian population is to benefit.

The apparatus called GENius condenses humid air on a cool core, filled with artificial leaves of plastic. It seems like an easy thing to do because we see it happening every day in summer with air condition but the tricky thing is to do it efficiently. Approximately 300 watt hours of electrical energy per liter of water produced are required. Water-Gen also uses the same process in dehumidifiers and tumble dryers.

GENius is supplied in several sizes – for schools, hospitals, villages or only for one family. It is still being tested in cities such as Mumbai and Mexico City before the launch in late 2017.

Nintendo and the unbeloved smartphone

Reading Time: 2 minutes

If there is something like a traditional company in the virtual world, it must be Nintendo. The Japanese giant, that just landed a coup with the augmented reality version of Pokémon, has been in the video games market since the 70s and developed countless classics from Super Mario to Zelda.

And its tradition does not only base itself on a rich history but also on the decisions made nowadays: Until today – Thursday the 15th of December 2016 – good old Mario has never run across the iPhone screen with Nintendo’s permission.

It is soon a decade ago, that Apple laid the cornerstone for today’s smartphone market. However, Nintendo has been relentlessly refusing to bring its popular game characters to the small computers. The Japanese hoped for the attraction of Super Mario, Zelda and Co. – who wants to play it, must buy a Nintendo console. In addition, they would get the sale of a console game equal to 40 to 60 euros or dollars.

In the meanwhile, the “free-to-play” model has been introduced to smartphone gaming, where the game itself is free and the providers then try to earn the money via virtual products or help to jump levels. A few games like Candy Crush collect millions, for many others it is a tough business.

The customers, however, definitely like to play on the smartphone – and Nintendo is now experiencing a steep fall in sales of its consoles and games. That’s why entering the business with mobile games has become almost inevitable.

Nintendo still wants to enter the market according to its own rules though. Only an initial level of Super Mario Run is free, sort of a demo. If you want to play the game completely, you will be charged 10 euros or dollars. This is much more than is common today with a mobile phone game.

Nintendo remains a renegade in this streamlined market – and already announces to tackle it with a new console, Switch, which will try to be a mobile and home device at the same time. The Japanese continue to jump and run between outside innovation and own tradition and seem to be more innovative than most while doing so.

Lost in translation no more? Google’s AI invents own language

Reading Time: 2 minutes

Google Translate has in all of its ten years of existence been a quite useful tool – appreciated by some (especially those who know about the difficulty of machine based translation), ridiculed by others, used by most. It has though never really lived up to professional standards. And, to take that upfront, that hasn’t changed since you last used it half an hour ago. But Google is about to use its main advantage. Its data. Its 140 billion translated words daily.

In order to do that, the ‘machine’ had to be made adaptive. One main obstacle down that road was that, considering 103 supported languages, the idea of creating comprehensive language pairs (meaning direct translation from each one language to possibly every other supported one), was close to impossible. The translation would usually go via English. And thereby any real progress was hampered because even if you managed to improve the result, what you would have improved would not have been the link between the two target languages but between either or both of those and English. As soon as you took English out of the equation, the progress would probably be lost. Translation is a delicate field, if you aim for good results.

What you want to do instead, according to American and German researchers, is creating a so-called neural network. A new common language that may serve as a link between languages that the machine has not been trained to directly translate. If it was able to translate between Hindi and Hebrew and Hebrew and French before, then it can now also translate between Hindi and French without any middle step. Another reason why neural network systems have hailed as the new star in the machine translation universe is that they take into account contextual meaning. Before, and still to an extent because all machine based translation by definition has to be statistical if doesn’t want to be random, a sentence was translated by translating the single words or idioms individually and then putting them together. This is how things like this happen:

55213b421462dd9658aa714da9e8534b

 

This brings us to the main point: Google’s new common language (or any neural network language for that matter), other than English, can be changed, improved, adjusted with the help of the users’ requests.

Which tempts some to the conclusion that for one of the first times, AI, Artificial Intelligence, is actually at play in a user-related field. But that’s a topic for another day …

Tagged , ,