IBM and ABB partner to focus on Internet of Things and Industrial artificial intelligence

ABB has sealed a collaboration agreement with International Business Machines Corp, the Swiss engineering company said on Tuesday, the latest step in its efforts to ramp up its presence in digital technology and the Internet of Things.

In a joint statement ABB said it would combine its digital offering, which gathers information from machinery, with IBM’s expertise in artificial intelligence featured in its Watson data analytics software. The two companies will jointly develop and sell new products.

“This powerful combination marks truly the next level of industrial technology, moving beyond current connected systems that simply gather data, to industrial operations and machines that use data to sense, analyse, optimise and take actions that drive greater uptime, speed and yield for industrial customers,” ABB Chief Executive Ulrich Spiesshofer said in a statement.

For example, instead of manual machinery inspections, ABB and IBM intend to use Watson’s artificial intelligence to help find defects via real-time images collected by an ABB system, and then analysed using IBM Watson.

ABB has identified digital technology – where machinery communicates with control centres to increase productivity and reduce downtime – as a driver of growth. It now gets around 55 percent of sales from products that are digitally enabled.

As part of the drive, the company last year signed a strategic partnership with Microsoft to roll out digital products for customers in the robotics, marine and ports, electric vehicles and renewable energy sectors.

To spearhead its strategy, it appointed former Cisco executive Guido Jouret as its first chief digital officer last year.

Reuters

Publish date: April 25, 2017 12:30 pm| Modified date: April 25, 2017 12:30 pm

Google’s new machine learning model can harness the processing power of your phone to improve AI

Image: Google

Google’s new machine learning algorithm harnesses the processing power of your smartphone to improve algorithms and save on data consumption by intelligent virtual assistants and the like.

Using a process called Federated Learning (FL), Google claims that they can “collaboratively share a prediction model while keeping training data on the device.” It might sound complicated, but it’s actually a simple, well thought-out solution to a problem that’s been haunting machine learning for years.

 Follow

sundarpichai 

✔@sundarpichai

I am excited about the progress here, really important work on machine learning without centralized training datahttps://research.googleblog.com/2017/04/federated-learning-collaborative.html …

6:05 AM - 12 Apr 2017

Federated Learning: Collaborative Machine Learning without Centralized Training Data

Posted by Brendan McMahan and Daniel Ramage, Research Scientists Standard machine learning approaches require centralizing the training da...

research.googleblog.com

 
•  213213 Retweets
 
•  659659 likes

Machine Learning (ML) is used to design algorithms. These algorithms help in the functioning of what we know as artificial intelligence (AI). Siri, Google Assistant, Bixby, they’re all just algorithms.

But ML is used for more than just digital assistants. Google’s self-driving car, for example, learns to drive using ML. Essentially, data is harvested and analysed and then an algorithm is developed based on the results. This one-line description grossly oversimplifies how ML works, but you get the general idea.

For example, if you normally drink your coffee at 6 AM every day on weekdays and at 8 AM on weekends, ML will analyse the data and determine your coffee drinking patterns, potentially automating your coffee making process if you have, say, a smart coffee maker. More data might lead the algorithm to determine that you drink coffee late on public holidays as well, or the data might be tied to your heart rate, etc.

The possibilities are endless, but are predicated on a large amount of data, some of it very revealing and potentially a privacy nightmare.

Image: Google

Traditionally, this data would need to be harvested by some device, like your smartphone, and then sent to the cloud for processing. ML is then used to analyse the data further and learn from it, further improving algorithms.

As Google explains in its blog-post, this approach requires low-latency, high-bandwidth data transfer rates. Of course, the privacy issues also raise their head here and taking responsibility for the data can put a company in a tricky spot.

With FL, Google has changed how the learning process works. As they describe it, “your device downloads the current model, improves it by learning from data on your phone, and then summarises the changes as a small focused update.”

The data that’s actually sent to the cloud is the updated algorithm and not the actual data itself. The data remains on your phone at all times.

Google’s cloud-based FL then aggregates these updates and creates an improved model for all devices. This updated model is then sent to devices as an update and the process repeats itself.

Using this technique, Google claims that bandwidth needs can be dropped by up to 100x.

Better yet, for the sake of privacy, the cloud-based processing won’t happen unless “100s or 1000s of users have participated.” This ensures that the updates from a single user cannot be identified.

The updated algorithms that your phone creates are sent to the cloud in a compressed and encrypted manner to begin with.

Google has deployed this technique in their Gboard app to help improve its natural language processing, but as they refine the process, they hope to expand FL’s scope to other applications as well.

FL has the potential to fundamentally change the way models are developed in the future. More information can be found here.

Publish date: April 12, 2017 11:16 am| Modified date: April 12, 2017 11:16 am

Nokia’s ‘Viki’ digital assistant will take on Google Assistant, Siri and Alexa in the near future

Image: Weibo

By tech2 News Staff /  09 Jan 2017 , 12:35

It certainly looks like the future of the smartphone is the digital assistant. The space is currently ruled by Google Now and Siri. Alexa and Cortana aren’t far behind either. Joining the fray, possibly, is Nokia’s own assistant and this one’s called Viki.

Google Assistant is an evolution of Google Now, but isn’t available on many devices yet.

GSMInfo.nl spotted Viki in a trademark filing made by Nokia. The trademark application was for the name Viki, and describes a software for “the creation and monitoring of mobile and web assistants working with digital knowledge and combining all data sources into a single chat and voice-based interface.”

Nokia, which recently revived its brand with the launch of the Android powered Nokia 6 in China, seems to have its heart in the right place. Samsung is also aware of the impending, AI-powered future and has been working on its own digital assistants for a while now (dubbed Bixby and Kestra).

As far as we’re aware, Viki is still only a name on a trademark application and we don’t even know the work that has been done on this AI.

There is also no timeline on if and when Viki will make its way to Nokia’s devices.

Microsoft opens up Cortana to developers and manufacturers with Skills Kit and Devices SDK

By tech2 News Staff /  14 Dec 2016 , 17:39

Microsoft is extending the capabilities of its artificial intelligence assistant, Cortana to new devices and developers. Microsoft announced a Skills Kit and a Cortana Devices SDK. The Skills Kit is to add capabilities to Cortana across the platforms that it is available on, including Windows, iOS and Android. The devices SDK will allow Cortana integration into a new generation of smart devices. Developers and Device Manufacturers can sign up to get access and updates as Microsoft moves the capabilities out of private preview.

The Cortana Skills Kit allows users to integrate capabilities with Cortana using chatbots created for the Microsoft Bot Framework. The web services offered by companies can be published as a skill on Cortana through the Skills Kit. There is also a facility to re-purpose existing code for Amazon’s Alexa. Early partners have already integrated skills into Cortana. These include Knowmail that uses Cortana to prioritise emails based on habits of individual users, Capital One that facilitates money management through a voice interface, Expedia that allows users to book hotels, and TalkLocal that can connect to local services based on user requests.

The voice based services can potentially be used by the premium smart speaker by Harman Kardon with Cortana Integration, that is expected to be available in 2017.

The Cortana Devices SDK will allow for third party Original Equipment Manufacturers and Original Device Manufacturers to integrate Cortana into their product offerings. The Devices SDK is platform and device independent, designed to make Cortana unbound, and work on any platform or device. Microsoft is already working with many industries across various hardware categories. One exciting tease from Microsoft is that it is working to get Cortana on at least one connected car. Cortana is included in the Core edition of the Windows IoT platform, which allows the assistant to be embedded in IoT devices.

The Cortana Skills Kit and the Cortana Devices SDK are in private preview as of now, and will be available more broadly in 2017.

IBM beta tests AI cyber security services from Watson for 40 global clients

Image Credits: REUTERS

By tech2 News Staff /  08 Dec 2016 , 09:32

IBM announced that more than 40 global clients in industries as varied as financial services, healthcare and education have joined the IBM Watson for Cyber Security beta program. The program uses the Watson AI to provide security services to these companies. Clients include Sun Life Financial, SCANA, California Polytechnic State University and Avnet.

The AI will help to identify and prioritise threats, allowing security analysts to make better and faster decisions. Watson uses machine learning and natural language processing to pull out the relevant information from vast amounts of data. A poll conducted by IBM Institute for Business Value among security professionals shows that more than sixty percent believe that emerging cognitive technologies will have an important role to play in combating cyber crime.

Sandy Bird, Chief Technology Officer, IBM Security, says “Customers are in the early stages of implementing cognitive security technologies. Our research suggests this adoption will increase three fold over the next three years, as tools like Watson for Cyber Security mature and become pervasive in security operations centers. Currently, only seven percent of security professionals claim to be using cognitive solutions.”

IBM Watson can determine if a security event or offence matches with the behavior of any known malware or cybercrime campaign. If so, then Watson can pull the relevant background information, and known approaches for tackling the malware or attack. IBM Watson also helps at identifying suspicious behavior, by providing contextual information about the behavior which will allow a security analyst to better decide if the suspect activity is indeed malicious.

IBM security monitors over thirty five billion security events every day, in more than a hundred and thirty countries, and IBM owns over three thousand security related patents.

Amazon goes all-in on AI and big data at AWS re:Invent 2016

Invent 2016 AWS

On Wednesday, at the first keynote of the AWS re:Invent conference in Las Vegas, Amazon Web Services (AWS) CEO Andy Jassy took the stage to explain a host of new updates to the cloud provider's portfolio of services. And, it seems Amazon is making a big bet on next -generation technology.

Some of the biggest announcements were the first three services of the Amazon AI portfolio. For starters, Amazon Rekognition provides image recognition, categorization, and facial analysis in batch analysis or real time. The facial analysis can detect sentiment, and tell whether or not the subject is wearing glasses, for example.

Amazon Polly is a text-to-speech (TTS) service that is powered by deep learning. It takes a text input and returns an MP3 stream that is altered to sound more like actual conversation. For example, if the text contains "WA" the output might say Washington instead. Jassy also announced that a new service called Amazon Lex, which powers Alexa, is coming as well. Lex provides natural language understanding and automatic speech recognition.


The new AI tools announced by Jassy could make it much easier for enterprise customers to tap into, and leverage, technologies such as machine learning to build their next generation of applications.


In his address, Jassy also noted that Amazon would be launching a new analytics product called Amazon Athena. As a companion to the existing EMR and Redshift, Athena is an interactive query service that allows users to analyze data in S3 using SQL. This significantly lowers the bar for everyday IT to utilize big data analytics to glean insights.

To go along with these new services, Jassy also announced a slew of new compute instances and features as well. Here are the following compute types that were announced:
T2.xlarge - 16 GiB, 2 vCPU
T2.2xlarge - 32 GiB,. 2 vCPU
R4 - 48 GiB, DDR4, L3 cache, 64 vCPUs
I3 - 3.3 million IOPS, 488 GiB, 15.2 TB NVMe SSD, 64 vCPUS
C5 - 72 vCPUs, Intel Skylake, 12 Gbps to EBS, 144 GiB


The C5 instance shows just how much hardcore GPU processing is being done on AWS. However, to give a broader group of users access to GPU power, Jassy announced a new service called Elastic GPUs for EC2, which allows users to attach a GPU to any of the existing compute instances in AWS.


Continuing in the vein of simplifying some of the features available in AWS, Amazon Lightsail was revealed as a way to make virtual private servers (VPS) easier to launch. Users choose from five bundles, name their server, and create it. Additionally, packages start at only $5 a month.

AWS users will also get access to a preview of F1 instances, Amazon's new FPGA instance family. This will allow users to run custom logic on EC2.

Additionally, to touch on IoT deployments, Jassy also announced AWS Greengrass. This service embeds AWS Lambda compute and other AWS services in connected devices, and allows users to manage them from the AWS console.

Last year, AWS launched Snowball, a secure appliance that makes it easier to move data to the cloud. At the 2016 re:invent, Jassy unveiled the general availability of Snowball Edge, which has on-board compute, and more storage than the previous version.
The 3 big takeaways for TechRepublic readers


Amazon announced three new AI products for image recognition, text-to-speech, and the natural language understanding that powers Amazon Alexa.
Amazon also announced Amazon Athena, an analytics products that allows users to query S3 with simple SQL.
The majority of the compute instance lineup in AWS got an update, with new products as well.

Amazon announces three new AI services called Lex, Polly and Rekognition for AWS

AWS

By tech2 News Staff / 01 Dec 2016 , 13:24


Amazon has started to offer artificial intelligence based services on its AWS platform, to give developers more tools to engage with customers. The three new AI tools are called Lex, Polly and Rekognition. Lex is the technology that powers Amazon Alexa, and allows developers to integrate rich conversational experiences in their offerings. Polly is a state of the art text to speech service that has forty seven life like voices in twenty languages. Rekognition is an image processing service, that can identify content in images.

Raju Gulabani, VP, Databases, Analytics, and AI, AWS said “The combination of better algorithms and broad access to massive amounts of data and cost-effective computing power provided by the cloud is making AI a reality for application developers. We are excited to see how customers use Amazon Lex, Amazon Polly, and Amazon Rekognition to build a new generation of apps that have human-like intelligence and can see, hear, speak, and interact with people and their environments”

Lex

Lex is the machine learning technologies that powers Amazon Alexa, with the key components being Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU). Lex can be used by developers to quickly make chat and voice bots, that can be integrated into services and applications. Lex is deployed as a fully managed service, requiring little time to set up, manage and scale.

The main concepts used by Amazon Lex. Image: Amazon.

Lex has an itnegrated development environment in a console. Developers can create bots, test them, and deploy them through the interface. There are some sample bots to start with as well. Bots built using Lex can be used on multiple platforms, and Amazon handles the authentication processes for different platforms. Lex can connect with Facebook Messenger as of now, but support for Slack and Twilio is being worked on. The Lex service is charged at the rate of $4 for 1000 speech requests, and $0.75 for 1000 text requests.

The testing interface for bots. Image: Amazon.

Benjamin Stein, Director of Messaging Products, Twilio said “Developers and businesses use Twilio to build apps that can communicate with customers in virtually every corner of the world. Amazon Lex will provide developers with an easy-to-use modular architecture and comprehensive APIs to enable building and deploying conversational bots on mobile platforms. We look forward to seeing what our customers build using Twilio and Amazon Lex.”

Polly

Polly is a cloud based text to speech service that generates human like voice based on a text string. The files can be downloaded as an mp3 for use in applications and services. Speech Synthesis Markup Language (SSML) is supported for advanced functionality, such as mixed language text. Developers can use SSML to indicate to Polly that some words in an English sentence are in French. There is a vast language end region menu, with support for five regional accents for English, including Indian. There are two alternative accents each for French, Portuguese and Spanish.

Polly supports plain text or SSML. Image: Amazon.

The SDK or console can be used to send text to Polly, which then converts it to speech in the cloud and beams it back. The service can be integrated into e-book readers, personal assistants, entertainment apps, public service announcement systems, or e-learning platforms. Polly can handle high volumes of text rapidly as well. Polly can return input over a command line interface as well.

some of the accents and languages available in Polly. Image: Amazon.

Joseph Price, Senior Product Manager, The Washington Post said “We’ve long been interested in providing audio versions of our stories, but have found that existing text-to-speech solutions are not cost-effective for the speech quality they offer. With the arrival of Amazon Polly and its high-quality voices, we look forward to offering readers more rich and versatile ways to experience our content.”

Rekognition

Rekognition is an image analysis artificial intelligence. Rekognition can be used to recognise faces, objects and scenes in an image. The AI delivers a confidence score for each identification, which is a rating of how accurate the identification is likely to be. These confidence scores can be further processed by an app or a service. There are advanced facial analysis functionalities such as face comparison, and face search.

Rekognition gives scores on a picture, confidently identifying the image as that of a “dog” and a “pet”. Image: Amazon.

Some of the capabilities of Rekognition include assessing if the mouth of a person is open or shut, whether or not they are smiling, if they are happy, whether they are wearing sunglasses and identify the presence or lack of facial hair. Applications for Rekognition include security services, smart marketing implementations that track user engagements, or automatic indexing and tagging for vast image libraries.

AWS has advanced features for processing images of faces. Image: Amazon.

Don MacAskill, Co-Founder, Chief Executive Officer, and Chief Geek, SmugMug said “SmugMug customers want to spend their time making more memories, not manually managing their photo collection. Amazon Rekognition will allow us to automatically identify the content in customers’ photos, unlocking a host of features that will allow them and their visitors to have more time to focus on enjoying life and celebrating their photos.”

The three new AI services are scalable and cost effective, with developers paying for only what they use. Amazon has simplified access to neural networks, data required for training, and expertise in machine learning. The heavy lifting is done by Amazon already, with the artificial intelligences trained for a wide variety of scenarios. Developers can directly start using the AI without building machine learning algorithms, training the AI with models, or commit to infrastructure investments up front.

Japan working on a 130 petaflops supercomputer, which could be the fastest in the world

Image: Reuters

  26 Nov 2016 , 11:18

Japan plans to build the world’s fastest-known supercomputer in a bid to arm the country’s manufacturers with a platform for research that could help them develop and improve driverless cars, robotics and medical diagnostics.

The Ministry of Economy, Trade and Industry will spend 19.5 billion yen ($173 million) on the previously unreported project, a budget breakdown shows, as part of a government policy to get back Japan’s mojo in the world of technology. The country has lost its edge in many electronic fields amid intensifying competition from South Korea and China, home to the world’s current best-performing machine.

In a move that is expected to vault Japan to the top of the supercomputing heap, its engineers will be tasked with building a machine that can make 130 quadrillion calculations per second – or 130 petaflops in scientific parlance – as early as next year, sources involved in the project told Reuters.

At that speed, Japan’s computer would be ahead of China’s Sunway Taihulight that is capable of 93 petaflops.

“As far as we know, there is nothing out there that is as fast,” said Satoshi Sekiguchi, a director general at Japan’s ‎National Institute of Advanced Industrial Science and Technology, where the computer will be built.

The push to return to the vanguard comes at a time of growing nostalgia for the heyday of Japan’s technological prowess, which has dwindled since China overtook it as the world’s second-biggest economy.

Prime Minister Shinzo Abe has called for companies, bureaucrats and the political class to work more closely together so Japan can win in robotics, batteries, renewable energy and other new and growing markets.

DEEP LEARNING

In the area of supercomputing, Japan’s aim is to use ultra-fast calculations to accelerate advances in artificial intelligence (AI), such as “deep learning” technology that works off algorithms which mimic the human brain’s neural pathways, to help computers perform new tasks and analyze scores of data.

Recent achievements in this area have come from Google’s DeepMind AI program, AlphaGo, which in March beat South Korean professional Lee Seedol in the ancient board game of Go.

Applications include helping companies improve driverless vehicles by allowing them to analyse huge troves of visual traffic data, or it could help factories improve automation.

China uses the Sunway Taihulight for weather forecasting, pharmaceutical research, industrial design, among other things.

Japan’s new supercomputer could help tap medical records to develop new services and applications, Sekiguchi said.

The supercomputer will be made available for a fee to Japan’s corporations, who now outsource data crunching to foreign firms such as Google and Microsoft, Sekiguchi and others involved in the project said.

The new computer has been dubbed ABCI, an acronym for AI Bridging Cloud Infrastructure. Bidding for the project has begun and will close on Dec. 8.

Fujitsu Ltd, the builder of the fastest Japanese supercomputer to date – the Oakforest-PACS, capable of 13.6 petaflops, declined to say if it would bid for the project.

The company has, however, said it is keen to be involved in supercomputer development.

Reuters

Google uses machine learning to upscale and enhance low quality photos

Image: Google

By Aditya Madanapalle /  15 Nov 2016 , 18:36

Image upscaling is often made fun of in movies and television series, for extracting impossible levels of image information from low quality, blurry photos. Typically law enforcement agencies approach a technician with blurry images, and after a few keystrokes, a high resolution images suddenly emerges. Google has developed technology called Rapid and Accurate Image Super-Resolution (RAISR), that actually works similar to what is shown in this clip from CSI.

The technology is not just for law enforcement agencies. The rapid evolution of both cameras and screens means that all of us have collections of older low resolution photos that we look at in newer high resolution screens. While there are upscaling and enhancing techniques around, none of them can fill in the missing information. Mostly, these algorithms work by guessing pixels based on the pixels nearby, creating aliasing artifacts. Google has a machine learning approach that does not directly bring in the missing image information, but does preserve and build on the underlying structure.

Google RAISR in action. Image: Google.

The technique uses machine learning to train on two sets of images, a high resolution version, and a low resolution version. The machine generates a set of filters that compares the two versions, and then enhances the low resolution image with the filters. If a high resolution image is not available, the technique can still be used. The AI is trained on the upscaled image based on traditional linear upscaling techniques.

Image: Google

RAISR can be used on mobile phones as well, and processes the images 10 to 100 times faster than traditional methods. The enhanced images are less blurry than current approaches. The RAISR technique can be used to improve pinch to zoom functionality in mobile phones. Lower resolution images can be sent over communication services to conserve on bandwidth, and these images can be restored to the original quality by the recipient.

You can see more examples of RAISR in action here.