Cool robot hand learns as it goes

It’s a device that brings to mind the bodyless hand, Thing, from The AddamsFamily: a human-like robotic hand, engineered by scientists at the University ofWashington, that can learn on its own as it handles a specific task.

The hand has five fingers, tendons, joints, over a hundred sensors, and iscapable of moving faster than its human counterpart. In a video the universityreleased, the hand can be seen delicately rotating a tube full of coffee beans–an activity that the robot can improve iteratively, the university said.

“Hand manipulation is one of the hardest problems that roboticists have tosolve,” Vikash Kumar, a doctoral student at the University of Washington andthe lead author on a new paper about the robot hand, said in a statement. “A lotof robots today have pretty capable arms but the hand is as simple as a suctioncup or maybe a claw or a gripper.”

But besides the impressive dexterity of this robotic hand, which is powered by apneumatic system and cost about $300,000, machine learning algorithms allowit to learn along the way.

“What we are using is a universal approach that enables the robot to learn fromits own movements and requires no tweaking from us,” Emo Todorov, anassociate professor at the University of Washington, said.

The next step for the device, the university said, is for the robot hand to not justlearn while performing a known task on an object it has worked with before, butto figure out how to manipulate new objects or handle new situations.

Google I/O 2016 Preview: A Chrome/Android merger? Project Tango? VR?

Google I/O is Google’s annual launch party where the company shows off its biggest products andteaches developers how to be a part of the Google ecosystem. The event takes place from May 18 to20 at the Shoreline Amphitheatre in Mountain View, California.

Google I/O this year is definitely weird. It’s in an all-new location, and the big news that we wouldexpect to happen at I/O—a new version of Android—has already been released. As a result, there isreally nothing that we would put at a 100 percent lock for I/O—we just aren’t sure what Google is goingto do. The schedule gives us some vague idea of what we can expect to see, but since Google likes toavoid spoiling its announcements, the schedule isn’t always a comprehensive list until after theopening-day keynote. While we aren’t making guaranteed predictions, we can at least offer a list ofpossibilities based on projects we know are in the works at Google HQ.

This is the first Google I/O since Google became a subsidiary under Alphabet, but we’re going toassume most of the other Alphabet companies aren’t going to join in on the fun. None of the divisions inAlphabet are really “new”—the groups have all been inside Google for some time. Even when theywere part of Google, most groups didn’t participate in Google I/O. Now that they’re separated evenfurther, we expect their non-participation in I/O to continue. The one exception is Google X (Now just”X”), which famously demoed Google Glass at the show in 2012.

After the N Developer Preview, what’s left for Android?

The Android L Developer Preview was announced at Google I/O 2014, and it immediately became thebiggest news of the show. The Android M Developer Preview was announced at Google I/O 2015, andstole the show there, too. The Android N Developer Preview was announced in early March with asurprise blog post. What does that leave for Google I/O 2016?

It’s a good bet a fresh version of the Android N Developer Preview will be released at the show, butthat’s usually not a huge deal—the second preview build introduced the Vulkan graphics API and a fewother features, and we wouldn’t expect the third build to be a much bigger deal. We’ve heard about newHTC-built Nexus devices in the works, but given that Android N has a Q3 2016 release date, I/O is tooearly to hear about those devices. We might hear about what sugary snack the “N” in “Android N”stands for. We’re hoping for more important news, though.

The great Chrome OS merger and desktop Android

The Wall Street Journal, TechCrunch, Business Insider, andRecode have all independently heard of a Chrome andAndroid merger that would result in Android becomingGoogle’s flagship desktop and laptop operating system. Themove would turn Android into Google’s “everything” OS,allowing it to run on watches, phones, tablets, laptops,desktops, TVs, and cars.

Android would need a brand new interface to work on a PCwith a mouse and keyboard, just as it has a brand newinterface for cars, TVs, and watches. We saw what seemslike the beginnings of this with Android N’s floating windowmode, a feature that is hidden and unfinished in the currentAndroid N Developer Preview. Perhaps Google I/O could be the coming out party for the multi-windowfeature or for the entire Android-on-PCs strategy.

The original Wall Street Journal report mentioned a release date of 2017, with “an early version”arriving in 2016. Google’s big demos usually happen at Google I/O, so could we see a demo of thedesktop version? Google would certainly want to get developer feedback on a major new interface likethis, and there would probably be lots of new APIs to talk about. It would also explain why we got arandom March release for Android N—Google didn’t want to overwhelm everyone by launching bothAndroid N and announcing a new desktop interface (and possibly a replacement for Chrome OS) atGoogle I/O.

Google Play comes to other desktop platforms?

Code inside Chrome OS suggests Google is planning to bring Google Play to its current desktopoperating system. The Google Play Store was being brought to Chrome OS as a Chrome extension viaGoogle’s “App Runtime for Chrome (ARC)” project. Chrome extensions run not just on Chrome OS, butalso on every desktop version of Chrome. If a Play Store on Chrome OS was released with this, itwould also be possible to have the Google Play Store run on Windows, Mac OS, and Linux. GooglePlay on other platforms is pure speculation on our part, but it would be technically possible, andit would line up with Google’s typical strategy.

On mobile, Google takes a “horizontal” strategy with iOS. It has its Android operating system, whichgets priority support for most Google apps, but it also makes 67 different apps for iOS. Wherever usersare, they can jump into the Google ecosystem. Android moving to the desktop would necessitateGoogle building desktop apps, and getting those apps on rival desktop OSes would mirrorGoogle’s mobile strategy. The Google Play Store on rival platforms would be an even more intenseplatform invasion. It doesn’t happen on iOS because the OS is strictly controlled by Apple, but thosecontrols don’t exist for Windows, OS X, and Linux.

Google would have a good pitch for developers, saying that Android code could be used on watches,phones, tablets, cars, televisions, the new Android desktop OS, Chrome OS, Windows, Mac OS, andLinux. It could also bootstrap developer interest for its fledgling Android-based desktop OS by tellingdevelopers they also get access to Chrome OS, Windows, Mac, and Linux users. Users that boughtgames or apps on one platform could bring them to all the others without a lot of extra work.

Google Play comes to China

With a population of 1.35 billion people, China is the world’s biggest market. It’s also a market thatGoogle Play doesn’t really do business in, thanks to past disagreements with the Chinese government.

Reports from several reliable outlets have said that Google is working with the Chinese governmentand plans to launch Google Play in China this year. According to Reuters, the plan involves a version ofthe Play Store “set up specifically for China, and not connected to overseas versions of GooglePlay.” There’s no telling if this will be announced at Google I/O, but being able to tell a room full ofdevelopers that they can now easily sell apps to an extra billion people seems like a win for Google.

Android is open source and available to anyone without restrictions, but the Google Play Store, GooglePlay Services (Google’s developer APIs), and other Google apps are not. To be a part of Google’secosystem, device manufacturers must sign an MADA (Mobile Application Distribution Agreement)contract with Google. When we last saw one of these confidential contracts in 2011, Google “bundled”all the Google apps together—if you wanted one you needed to include all the others. It also includedan anti-fragmentation clause that banned participants from taking “any actions that may cause or resultin the fragmentation of Android,” including creating or distributing “a software development kit (SDK)derived from Android or derived from Android Compatible Devices.” This is widely interpreted as “youaren’t allowed to use Android on any product without also including Google Play.”

Since Google has every major Android manufacturer under contract, in most countries Google Playacts as the “default” app store and has a huge market share of Android-based devices versus otherapp stores. China is the main exception. With no Google Play Store in China, there is no “default” appstore, so it’s a messy free-for-all. Chinese customers end up with an OEM-provided app store andmaybe one from one of the major Chinese Internet companies, like Tencent, Baidu, or Qihoo 360.

Every OEM in China is basically shipping “forked Android” without Google Play, which would seeminglyrun afoul of the anti-fragmentation wording. Remember we’re working off of older MADA contracts from2011, but back then the contracts contained an interesting loophole that allowed Chinese app stores toexist. The contract deals with distributing Google applications “in the territories specifically authorizedby Google.” China is not one of those countries, hence the app store free-for-all.

Google notes that this list of territories “may be updated by Google from time to time.” So our nextquestion is “Then what?” What happens when Google Play moves into China? Does everyone’s MADAcontract kick in and Xiaomi, Huawei, and other top Chinese OEMs all immediately have to startshipping Google Play on everything? The transition will be something to keep a close eye on.

Yet another reboot of Google’s instant messaging strategy

Many of the world’s biggest tech companies consider instant messaging to be an important businessworthy of their money and development resources. Snapchat and Facebook’s WhatsApp areboth valued at $16 billion. And that’s nothing compared to China instant messaging app WeChat, whichis valued at a whopping $83.6 billion. Judging by the products it has so far produced, it seems Googledoes not agree with this assessment of the messaging market. The company’s only IM product isGoogle Hangouts, a neglected, barely developed product that has earned an awful reputation withAndroid users.

About six months ago, The Wall Street Journal reported that Google is planning a “new, smartermessaging app” that presumably would replace Hangouts. The new messaging app would integrate”chatbots” that would answer questions from “the Web and other sources.” Microsoft is working on asimilar concept inside Skype. At Build 2016, the company demoed the ability to do things like orderpizza from Domino’s via a chatbot with automatic escalation to a human.

A “visual search” camera app

Google Goggles is a little-known visual search app developed by Google. You snap a picture, it getsuploaded to Google’s servers, and Google tries to figure out what’s in the picture. It can identifyartwork, QR codes, and landmarks. The app can also solve Sudoku puzzles and do other things thatwould be difficult to search for with purely text.

The standalone app in the Play Store is pretty neglected, but a report from Slashgear says that Googleis planning to revamp its visual search offering by integrating it directly into the Google Camera, thedefault camera app on Nexus devices. The report says users would be able to snap a picture and tracearound the item they want more information on, and Google would return with results.

Apple Watch 2 News: Rumors, Price, Specs and Release Date Update

He stated that the Watch will keep getting better and that the company is still learning. According to him, in a few years time, people will wonder why they didn’t get the watch before.

“You’ll see the Apple Watch getting better and better. We’re still in learning mode. We’re learning fairly quickly, though. We know a lot more than we did a year ago.”

“I think that in a few years we will look back and people will say, ‘How could I have ever thought about not wearing this watch?”

Update – April 27, 2016 – A new report suggests that the next generation Apple Watch will feature true cellular connectivity. This means that you’ll be able to make calls, send texts and connect to the internet without the need of your iPhone.

Few smartwatches from LG and Samsung already offer 3G or 4G connectivity. Therefore, it would make sense for Apple to include the feature in the Apple Watch 2. The Wall Street Journal has heard from sources that Apple is working on bringing cellular connectivity and a faster processor to the next Watch.

Update – April 14, 2016 – The Apple Watch 2 may not do as well as Apple hopes, says Ming-Chi Kuo of KGI Securities. He has been known to have a good track record when it comes to Apple hardware and has been right several times.

In his latest research note, he states that Apple Watch sales will decline in 2016, despite the launch of a new updated device. One of the main reasons for this, is that the Apple Watch 2 will not see any design changes. Instead, there will be updated internals, a possibly thinner design, improved battery, and a FaceTime camera. It is also rumored to work without an iPhone.

He believes that the Apple Watch 2 will begin production in quarter 3 of 2016 and will launch with the iPhone 7 in September.

Another report that came out last week stated that the next-gen Apple Watch will be at least 40 percent slimmer. We’re not sure if Apple can managed that without affecting battery life.

March 18, 2016 – Apple will be unveiling new Apple Watch bands on March 21. The company sent out invites for the special event earlier this month. We also expect to see the all new 4 inch iPhone SE, new iPad Air 3 or iPad Pro mini, updates to Macs, iOS 9.3 and watchOS 2.2.

Recent rumors still suggest that the Apple Watch 2 will be released in Q2 2016. According to the Chairman of Quanta, Barry Lam, the next-gen Apple Watch will be unveiled towards the end of Q2 2016.

Quanta was the company that help Apple manufacture the first gen Watch. We believe that Apple will launch the Apple Watch 2 in September 2016.

Update -March 06, 2016 – It still looks like that Apple will not be releasing a new improved Apple Watch 2 with upgraded internals and a FaceTime camera anytime soon.

We may only see the new smartwatch in September this year, with Apple following yearly release cycle. Reports did suggest that Apple will instead launch an updated Watch model with newer straps and software.

Update – February 8, 2016

Recent rumors suggest that Apple will not be releasing the Apple Watch 2 in March this year. Instead, the current gen models will receive new bands and watchOS 2.2. That’s what’s going to happen.

However, several online retailers and brick-and-mortar stores are selling the Apple Watch at significant discounts. Are they getting rid of their inventory to make space for new ones? It could be that these retailers are preparing for new shipments of an updated Apple Watch, instead of a new model.

British online retailer Currys PC World is giving £50-120 discounts on the Apple Watch. We’re not sure if this means that they’re making room for the Apple Watch 2, or just trying to have a good sale.

A March 15 event is said to unveil the rumored iPhone 5se, the new Apple Watch updates, iPad Air 3 and other software upgrades. Let’s see what Apple has to offer.

Update – January 23, 2016

The second gen Apple Watch 2 will not be coming in March this year, according to a new report from TechCrunch.

Matthew Panzarino from TechCrunch has heard from several sources that Apple may not unveil a new Apple Watch model in March.

Instead, a slightly improved version of a FaceTime camera is what Apple will announce this year. There won’t be any changes to the design, case or straps.

I’VE NOW HEARD A BIT MORE THAT SUGGESTS THAT APPLE MIGHT SHIP A MINOR REVISION OF THE APPLE WATCH THAT INCLUDES A FACETIME CAMERA AND NOT MUCH ELSE — BUT STILL THAT IT WOULD NOT BE A FULL “WATCH 2.0” WITH CASING CHANGES AND MAJOR IMPROVEMENTS.

 

Previous rumors had suggested that Apple may be having a special event in the month of March where it would unveil the Apple Watch 2. Supply chains had also mentioned a March 2016 release date, but it looks like that may not be the case. The latest gossip from the Supply chains is that there isn’t anything substantial going on.

Apple Watch 2 Release Date

Apple should have started production in late 2015 if it had planned to release the new Apple Watch in March 2016. Yes, there have been rumors stating that Apple has indeed started production, but that may be entirely correct.

There haven’t been any chatter regarding new software for the second gen Apple Watch. Although Apple has been releasing updates to watchOS 2, there hasn’t been any significant update yet. As per the reports, the second-gen smartwatch will be a separate piece of hardware. It will have its GSM antenna, better battery life, faster hardware and maybe a slight design refresh.

The Apple Watch 2 will come out this year, but it may not be released in March. The report says that a September release alongside the introduction of iPhone 7 and iPhone 7 Plus is more plausible. Apple might instead refresh the iPads and release the rumored 4 inch iPhone in March.

Apple Watch 2 Rumors and Release Date: May Not Be in March

Five things to know about using Apple Pay in Canada

BUILD YOUR OWN RASPBERRY PI TERRARIUM CONTROLLER

Tom Bennet grows Nepenthes, tropical carnivorous plants that I know by the nameof pitcher plants. To stay healthy they need a temperature- and humidity-controlledenvironment, and Tom ensures this by housing them in a terrarium controlled by aRaspberry Pi 3 and Energenie’s Pi-Mote starter kit, which provides an easy way tocontrol mains electrical sockets from a Pi. He has written step-by-step instructionsto help you build your own terrarium controller, the first such guide we’ve seen forthis particular application.

Tom’s terrarium controller doesn’t only monitor and regulate temperature, humidityand light, three of the four main variables in a terrarium (the fourth, he explains, iswater, and because terrariums tend to be nearly or completely sealed, this requiresonly infrequent intervention). It also logs data from its sensors to Internet-of-Thingsdata platform ThingSpeak, which offers real-time data visualisation and alerts.

One of the appealing aspects of this project, as Tom observes, is its capacity forextension. You could quite easily add a soil moisture sensor or, particularly for aterrarium that houses reptiles rather than plants, a camera module, as well asusing the online data logs in all kinds of ways.

The very clear instructions include a full and costed bill of materials consisting ofoff-the-shelf parts that come to less than £90/$125 including the Pi. There arehelpful photographs and wiring diagrams, straightforward explanations, practicaladvice, and Python scripts that can easily be adapted to meet the demands ofdifferent habitats and ambient conditions. Thank you for writing such a usefulguide, Tom; we’re certain it will help plenty of other people set up their own Pi-controlled terrariums!

Neutrons tap into magnetism in topological insulators at high temperatures

May 9, 2016 by Jeremy Rumsey

A multi-institutional team of researchers has discovered novel magnetic behavior on the surface of a specialized material that holds promise for smaller, more efficient devices and other advanced technology.

Researchers at the Department of Energy’s Oak Ridge National Laboratory, Massachusetts Institute of Technology and their collaborators used neutron scattering to reveal magnetic moments in hybrid topological insulator (TI) materials at room temperature, hundreds of degrees Fahrenheit warmer than the extreme sub-zero cold where the properties are expected to occur.

The discovery promises new opportunities for next-generation electronic and spintronic devices such as improved transistors and quantum computing technologies. Their research is discussed in a paper published in the journal Nature.

TIs are relatively new materials, said Valeria Lauter, coauthor and instrument scientist at the Spallation Neutron Source, a DOE Office of Science User Facility at ORNL. A unique property of TIs is that electrons can flow on the surface without dissipation, while the bulk of the material serves as an electrical insulator—ideal for semiconducting materials.

“The properties of TIs are fantastic,” Lauter said, “but in order to use them in devices we need to be able to introduce magnetism on the surface without disturbing the bulk insulating properties of the material.”

This can be achieved in two ways: by impurity doping, where magnetic atoms are incorporated onto the TI surface, or by proximity coupling, where magnetism is induced by interfacing the TI with a layer of ferromagnetic insulating film.

The first method presents a problem, however. Doping can cause magnetic clusters if the atoms are not uniformly distributed, resulting in decreased electron transport controllability. Proximity coupling, on the other hand, can be obtained on clean, atomically sharp interfaces with crystalline orientations between two materials.

Using the proximity coupling method, Lauter’s collaborators at the Massachusetts Institute of Technology engineered hybrid bilayer heterostructures of bismuth selenide (Bi2Se3) TIs combined with a europium sulfide (EuS) ferromagnetic insulator (FMI). The definite spin directions of the FMI in proximity to the TI enable dissipation-free, spin-polarized (i.e. magnetic) electron flow in a thin layer close to the interface. That marriage forms a mutually magnetic relationship, Lauter said, though it’s difficult to establish.

“The first challenge is to grow the system,” she said. “The second is to measure its magnetism— not an easy thing to do when the small magnetic signals are hidden between two materials.”

The bulk EuS itself presents a particular challenge in that it’s limited by a low Curie temperature (Tc), the temperature at which a material ceases to demonstrate ferromagnetic behavior—in this case a temperature of approximately 17 kelvins (17 K), or negative 430 degrees Fahrenheit, well below a suitable room temperature for electronic devices.

To identify the hidden magnetic signals, Lauter used a polarized neutron reflectometry (PNR) technique at the Magnetism Reflectometer instrument on SNS beam line 4A. Neutrons are well-suited for this type of detection because of their sensitivity to magnetism and their innate ability to pass through materials in a nondestructive fashion, elucidating structural and magnetic depth profiles. Likewise, PNR is suited to studying Bi2Se3-EuS interfaces because it is the only technique that can measure the absolute value of magnetic moments in the materials.

The first sample measurements were taken at 5 K, well below the EuS Tc of 16.6 K. From there, Lauter took measurements above the Tc, starting at 25 K, and to her surprise, the system was still highly magnetic.

“This was quite unexpected. Above this temperature [16.6 K] nothing in the system should be magnetic,” Lauter said. “So I measured at 35 K, then 50 K, and it was still magnetic. I measured all the way up to room temperature [300 K, 80 F] at several points and saw that a small magnetization was still present.”

Lauter notes at room temperature the level of magnetism is reduced by more than a factor of 10 compared to its 5 K value. Nevertheless, she says, it remains substantial considering no magnetism is found in EuS layers above 50 K without the TI interface.

To substantiate the results, subsequent measurements were taken using different samples with varying thickness combinations. Throughout the experiments, neutrons revealed that ferromagnetism extends approximately 2 nanometers into the Bi2Se3 from the interface.

“This discovery could open new doors for designing spintronic devices,” Lauter said. “Ferromagnetic surface states in TIs are also thought to enable the emergence of exotic phenomena such as Majorana fermions—potential building blocks for quantum computers.

“These are just the properties we know of today, and we’re continuing to find even more.”

AI is ‘most important’ tech today, but it can be misused, says Microsoft exec

Artificial intelligence is “the most important technology that anybody on the planet is working on today,” observed Dave Coplin, the chief envisioning officer at Microsoft U.K. Coplin made that statement a room full of business leaders at an AI conference in London last week, reported Business Insider.

Companies are investing in AI left and right, and the tech is developing by leaps and bounds. Most people walk around with some form of AI in their pockets every day.

From the simplest applications, like Siri and Cortana, to complex driverless cars, AI has been doing more and more.

According to Coplin, however, it’s not what has already been achieved with AI, but what is possible in the future.

Coplin dug deeper, saying AI “will change how we relate to technology. It will change how we relate to each other. I would argue that it will even change how we perceive what it means to be human.”

Business Insider noted that AI has potential to aid in crucial scientific research such as cancer research.

Still, Coplin isn’t ignorant to the fact that what can be used to do good can, more often than not, also be used to do harm.

The tech is new, Coplin noted, so we don’t know what kind of biases people hold and how those biases will affect the AI they develop. “We’ve got to start to make some decisions about whether the right people are making these algorithms,” Coplin said.

Coplin isn’t the only tech giant to step in with some real talk about the dangers of AI. Elon Musk, CEO of Tesla and founder of the non-profit AI organization OpenAI, has spoken widely about the risks of AI.

In a past blog post about OpenAI, Musk wrote, “It’s hard to fathom how much human-level AI could benefit society, and it’s equally hard to imagine how much it could damage society if built or used incorrectly.”

Team Boosts Lithium-Ion Battery Performance

Google is testing a radical change by turning people’s search results black

https://vine.co/v/iXXQO1I0uIV/embed/simple