Weekly Digest #9: Facebook's AI, Healthcare IoT Growth and Smart Parking in Belgrade
June 7th, 2016
Last week, the analysts presented several reports regarding the development of mobile technologies, Internet of Things, and digital health while Facebook disclosed the details about their own artificial intelligence project.
The Growing Number of IoT devices
Ericsson presented their annual report called Ericsson Mobility Report where a lot of attention is paid to the Internet of Things (IoT) technologies. The report analyzes how mobile technology will be developed as a whole, but it is the first time a whole section has been devoted to IoT-devices. According to the information presented, the number of IoT-devices will exceed that of mobile phones by 2018. By 2021, there will be 28 billion connected gadgets worldwide, 16 billion of which will be IoT-gadgets.
By 2021, there will be 28 billion connected gadgets worldwide, 16 billion of which will be IoT-gadgets.
In terms of technology, the most actively growing number of connected devices are mobile gadgets (Cellular Internet of Things). Speaking geographically, the leader in volume of connections is Western Europe. Here, in five years, the number of connected devices is forecast to quadruple and the number of smartphones - to double. This year, the number of smartphones with an internet connection will outnumber simple phones.
As for the growth of the mobile traffic volumes, Asia-Pacific region is the leader, with Eastern Europe in the second place. Users are mostly consuming the mobile traffic and traffic generated by social networking users.
Healthcare Sector IoT Technologies
Another report forecasts active IoT-technologies growth, but for the healthcare sector. Grand View Research authors expect that this market sector will grow up to $410 billion in 2022. For comparison, in 2014 the market volume was $58.4 billion. Main drivers for the market will be technologies for medical equipment connection as well as software for analysis of data transmitted from the patients’ IoT-gadgets. An important factor in this growth will be patient’s medical data and analytical tools and technologies. Analysts expect that implanted medical equipment market will also increase, as well as ordinary consumer devices that analyze the person’s body state - from heart rate to blood sugar levels. Geographically, IoT in healthcare has developed the fastest in North America and in the upcoming years Asia-Pacific is expected to grow as actively.
Little IOT Agent
Several companies at once have announced their own developments in the field of Internet of Things. VMWare company, developer of instruments for visualizing, announced the launch of Little IoT Agent - a set of open source software development, created to help developers create secure IoT data and manage applications. The company claims that Liota can be used for app creation that can work through gateways from different vendors. Liota can work on any device that supports Python programming language. Bask Iyer, IT director, wrote in the company’s blog: “The Liota Open Source SDK provides the libraries to develop applications that connect and orchestrate data and control flows across things, gateways and the cloud.”
The Liota Open Source SDK provides the libraries to develop applications that connect and orchestrate data and control flows across things, gateways and the cloud. (Bask Iyer)
Cisco & Watson Collaboration
In the meantime, IBM and Cisco have joined forces to create analytical tools for the IoT. The agreement provides for shared usage of Cisco network devices and Watson analytical software. Because of this, various devices can be connected to this network and their work can be analyzed. The main peculiarity of working with Watson is that you can connect stand-alone devices as well as unmanned gadgets to this IoT network and the connection may even be automatic without human intervention.
Finding a Parking Place Becomes Easier
Internet of Things tools implementation can be absolutely different. It’s not just a fridge or a washing machine connected to the Internet, but also a “smart” parking. Any big city these days is struggling with the search of free parking places. In Serbia’s capital, Belgrade, a company called Parking Servis, which deals with paid parking, began implementing a system called Sensit.
The system informs the drivers about free parking spaces within city limits using infrared and magnetic sensors to determine whether there is a car in the parking space or not. The status of the place is transmitted to special digital tags next to the parking zone and broadcasted to inform about the availability of parking places in any given period of time.
The goal of the system is to provide a more uniform traffic, helping the drivers to pave their routes directly to areas with available parking spaces instead of “winding circles” in a search for one around the city. Having found a free space and parking there, the drivers can then pay for parking via mobile app provided by Parking Servis. At the initial stage, Sensit system will be connected to the 850 parking sensors on Njegoševa street. In the future, 3,500 sensors are planned to be spread out throughout the Belgrade. The app for drivers is based upon integration with the Nedap sensors network.
Facebook DeepText AI
Almost all IT-market leaders are announcing their own developments in the field of artificial intelligence. The time has come for Facebook to admit that too. Social network team has developed an AI system called DeepText AI. The main objective of this system is to improve the content personalization for users. DeepText AI will also help blind social media users to understand what is depicted in the photographs. In addition, it can help Facebook Messenger’s users to order a taxi, book tickets, and perform many other useful tasks (which is the most obvious way of monetizing DeepText AI). Moreover, the AI will understand questions like “I’d like to sell my old bike for $200, who is interested in such a proposal?” DeepText will also be used to automate the configuration required to conduct financial transactions on the Internet.
In the Facebook blog for developers, experts have explained that DeepText AI uses neural networks in combination with machine learning platforms, libraries such as FBLearner Flow and Torch.
“I’d like to sell my old bike for $200, who is interested in such a proposal?”
How Fast Is Your Website?
Google has introduced a tool to check the website loading speed and ease of viewing on mobile devices. The instrument was created by Think with Google division. To check your website, you only need to enter the website’s URL and the system will show you the statistics (and offer to send you a detailed report via e-mail).
We Don't Want Ads
In closing, one more interesting report of the last week - the analysis of ad blockers usage on mobile devices. The main conclusion of the report, in general, is obvious: more and more users install ad-blocking tools. According to PageFair analytical firm, the number of AdBlock-analogs has grown 90% over 2015. Today, 1 out of 5 smartphone owners blocks advertising. It is noteworthy that most active ad-blocking users come from Asia and especially China. The most popular mobile browser that has built-in ad-blocker is UC Browser.
