New Florence. New Renaissance.

IoT and Voice interfaces on the farm

Mrinal Wadhwa, CTO at the IoT company, Fybr recently shared with me this interesting use case in farming.

"A plum farm lost $ 2 million dollars in revenue, last year, because the heat affected the crop and truck loads of fruit were rejected by Costco.

Fybr is working with this plum grower to create an early warning system with infrared sensors that continuously monitor the amount of heat that the tree canopy is experiencing. In contrast to weather forecasts, this sensor information is timely, granular, local and actionable. A farm manager may turn on irrigation sprinklers to cool down the field on a particularly hot day. Fybr also retrofitted their irrigation pumps and drip lines for wireless remote control, a step towards automating the sense-and-respond process.

Low cost sensing, remotely controllable equipment and data driven learning algorithms will help farms in the near future protect against the risk of sudden weather events, significantly reduce wasted water and improve crop yields. As the amount of real-time information available about their farm grows, it's impractical for farmers to navigate through apps on a tablet or a phone while they work in their fields. Voice based command and query will be a potent tool for growers to proactively respond to changing conditions.

Older voice technologies, like Apple's Siri, had minimal customization capabilities for enterprises, but the latest with Google's Assistant/Dialogflow/Cloud Speech and Amazon's Alexa/Lex/Polly provide powerful tools for enterprises to quickly build high quality natural language interfaces.

In the below video, Fybr's Alexa integration is able to interpret and identify the correct irrigation pump and opens it successfully. Even a small error rate in this would be unacceptable in business critical applications. The next challenge - improving user identity and authentication around voice technologies."

 

Microsoft’s Chief Literary Officer

Cortana is Microsoft’s personal AI assistant for smartphones, named for a fictional character from the company’s sci-fi video game series Halo.Strictly speaking, engineers coded her into existence, but her well-timed wit comes courtesy of Microsoft’s resident blabbermouth, principal content publishing manager Jonathan Foster, and his creative posse of novelists, journalists, poets, filmmakers and playwrights who together shape the sass and spunk behind the gabby digital helper.

So the need to soften a bot’s cold, calculated edges has led to a new emerging path for poets, comedians and fiction writers like Foster inside tech companies. Foster, who earned an M.F.A. in film and playwriting from the University of Southern California, has written several plays and one novel — and loves writing witty banter.

Oxy

See how Amazon Siri, Google Assistant, Amazon Alexa and Microsoft Cortana would do in a stand up routine

 

Facebook’s Sounds of Silence

From a post by Regina Dugan, VP of Engineering at Facebook

“Today, at F8, we announced two projects focused on building new capabilities for communication. Think of them as “silent speech interfaces” with all the convenience of voice and the privacy of text. We asked these questions: What if you could type directly with your brain? And what if you could hear with your skin? The answers reinforce what we intuitively know…

We are wired to communicate. And connect.

Over the next 2 years, we will be building systems that demonstrate the capability to type at 100 wpm by decoding neural activity devoted to speech. Just as you take many photos and decide to share some of them, so too, you have many thoughts and decide to share some of them in the form of the spoken word. It is these words, words that you have already decided to send to the speech center of your brain, that we seek to turn into text. And unlike other approaches, ours will be focused on developing a non-invasive system that could one day become a speech prosthetic for people with communication disorders or a new means for input to AR. Even something as simple as a ‘yes/no’ brain click, or a ‘brain mouse’ would be transformative.

We also described a system that may one day allow you to hear through your skin. You have 2 square meters of skin on your body, packed with sensors, and wired to your brain. In the 19th century, Braille taught us that we could interpret small bumps on a surface as language. Since then many techniques have emerged that illustrate our brain’s ability to reconstruct language from components. Today we demonstrated an artificial cochlea of sorts and the beginnings of a new a ‘haptic vocabulary’.”

Photo Credit from F8

Translations at WEF at Davos

At the World Economic Forum in Davos, Switzerland, last week, Layer provided messaging tools as part of the Forum’s app that allowed one user to send a message in her native language – and another user to receive it in his native language.

“The attendees are obviously from all over the world, and they’re all seeking to understand one another, so being able to facilitate that communication is pretty powerful,” said Ron Palmeri, Layer’s chief executive.

To add the instant translation ability, Layer used the Microsoft Translator API, which lets developers build translation into their apps and other tools. Layer said it only took about four hours to add the translator to their messaging tool using Microsoft’s API, and Palmeri said they expect to use translation capabilities for other clients in the future.

Microsoft Blogs

The STEM angles in the movie Arrival

“I couldn’t tell if the actual movie would be mostly science fiction or mostly a love story.” says Stephen Wolfram (the creator of Mathematica) who along with his son has a consulting role in the movie.

I have seen the movie twice and cannot make up my mind either or indeed if it is more about a third topic - about communication between humans and with aliens.

The lead character played by Amy Adams is a linguist. We learn early on she is familiar with Portuguese, Farsi, Sanskrit and the movie weaves in trivia like Urdu is written from right to left.

So, what does that have to do with STEM?

To start with the movie is adapted from Story of Your Life, a short story by Ted Chiang. In linguistics, it's known as the Sapir-Whorf hypothesis, “or as Chiang puts it in the film's production notes "the idea that the language you speak determines how you perceive the world and even what kinds of thoughts you can have." It may even determine how your brain is wired. (As an aside, it is incredible that a short story has been adapted into such a visually dense movie, and with other references to Fermat’s Principle of Least Time and Bayes’ Theorem excised so the movie did not confuse the audience even more.)

And then there is the logogram language of the aliens. As Wired reports “The aliens regard time as non-linear, and the language needed to reflect that. But consultations with linguists and graphic designers kept leading to fictional alphabets that Vermette says hewed too closely to familiar systems like hieroglyphics, or code. It felt too human. Then one night, Vermette’s wife, artist Martine Bertrand, offered to sketch some ideas. The next morning, Vermette came downstairs to find 15 inky logograms on the kitchen table. “I said, ‘eureka.'””

Finally, here is repeated use of 12 – the symbol of cosmic order, as in 12 signs of the Zodiac, 12 year cycle in Asia. 12 spaceships show up in the movie, and there is science which tries to decode their landing sites. Jeremy Renner who plays a theoretical physicist uses gobs of computing power to analyze the alien logos into 12 sectors.

The non-linear time dimension in the movie is tough to get your head around. Very few of the alien logograms are actually translated into English sub-titles in the movie. Finally, bringing out the human elements which actually dominate the movie, pose an interesting challenge. As Jeremy Renner explains

“How I deal with, like, zeroes and ones, and [makes a sound like a chattering computer, or maybe an old dot matrix printer] and all these really unemotional, scientific things? How do we make this guy a human?”

It is a fascinating movie. Go see it – maybe 2-3 times to finally get your head and heart around all the nuances.

Photo Credit 

 

 

Smart Tattoos

Forget smartphones and watches. DuoSkin, a new product from Microsoft and the MIT Media Lab, can turn your epidermis into a touch pad. Or a remote. Made with naturally conductive gold leaf, DuoSkin places a technological interface directly on your body. One potential application would include a Bluetooth-like chip so wearers could sync their temporary tattoo to a music system, then swipe it to change volume. Another version changes colors. The process is designed so people can customize both form and function for a tattoo that lasts only a day. The goal, says lead researcher Cindy Hsin-Liu Kao, is to create technology that’s as personal as lotion or makeup, so “it really blends into the wearer’s identity.”

Time (sub required)

 

The versatile Apple Pencil

Despite Apple's claims to the contrary, not everyone believes the iPad Pro can replace a computer. When paired with an Apple Pencil, however, the tablet lets people do things they might never accomplish on a traditional PC or Mac. For example, the iPad Pro lets you create handwritten thank you cards that can then be sent through the mail, write notes on PDFs and sign on the dotted lines, and drain all the color out of photos, then add it back to particular objects.

The following 12 iOS apps look great on iPad Pro, and they all put your Apple Pencil to work.

 

 

The US Navy’s “Magic Carpet”

The autopilot tracks the position of the deck, adjusting the throttle, flaps, ailerons, and stabilizers to keep the flight path and angle of attack on point. Instead of maintaining continuous pressure on the stick and making myriad inputs before landing, the pilot can relax. Any adjustments he does make are incorporated into the autopilot settings.

During a week of trials last month, test pilots flying F/A-18 Super Hornets conducted nearly 600 touch-and-go landings and many tailhook-arrested landings on the Nimitz-class USS George Washington. They made both highly accurate approaches and deliberately inaccurate approaches, with varying wind speeds and directions. According to engineers with the Navy and Boeing, the system increased the accuracy and consistency of landings under all conditions. Those landings were less stressful, too: Pilots typically perform 300 corrections to their flight path in the final 18 seconds of an approach. Magic Carpet drops that between 10 and 20.

Wired

 

 

Machines and voice learning

(intelligent Voice)’s CEO Nigel Cannings says the breakthrough came when he decided to see what would happen if he pointed a machine-learning system at the waveform of the voice data – its pattern of spikes and troughs – rather than the audio recording directly. It worked brilliantly.

Training his system on this visual representation let him harness powerful existing techniques designed for image classification. “I built this dialect classification system based on pictures of the human voice,” he says.

New Scientist

 

New Frontiers for Apple UX

Wired on announcements this week

3D Touch - By weaving 3D Touch more deeply into the fabric of iOS 10—and specifically, making it an essential part of its lock screen—Apple finally arms potentially revolutionary feature with true purpose. You may not use 3D Touch much today, but after iOS 10 arrives this fall, you may wonder how you lived without it.

Siri - Not only that, you’ll be able to use Siri on your PC, to make a lot of simple actions easier: adding things to your calendar, doing quick research and calculations, setting reminders, playing music, even searching your computer. Siri can search Finder, finding you files from last week about the offsite and then showing you the ones you tagged as draft. Click on a button and it pins into your notification center, for easy finding later. The voice assistant can do more on the Apple TV as well: Siri has improved topical searches for movies and TV shows (“Horror movies from the ’80s”) and you can now run voice searches for YouTube videos.