Weekly Digest #8: Twitter Analytics and Internet of Things Network
May 31st, 2016
Samsung has announced last week that they are planning on creating a network for the Internet of Things, Twitter has given access to real-time analytics, and IDC forecasts a strong growth in demand for business analytics.
National IoT Network
South Korea is one of the most advanced countries in the world. It has been the flagship in IoT-gadgets development and introduction of the “Smart Home” technologies, and now they will be the first to create a dedicated network for such gadgets and sensors. South Korean tech giant Samsung Electronics and mobile operator SK Telecom plan to launch a national network that brings together all IoT devices in South Korea. The goal of such network is to become a supplement to existing solutions and it will be based on LoRaWAN technology. LoRaWAN (global computer long-range network) provides transfer of small amounts of data over long distances. Such network consumes very little power but provides fast data transfer. Because of this, it will be an ideal solution for linking smart devices’ sensors. The network will launch in Daegu, a South Korean city. The system testing will include its performance, possibility of collecting various data, its analysis and forecasting (for example, weather, traffic, and air pollution information analysis.)
New Types of Memory
The IoT concept poses new challenges to the usual data exchange and storage technologies. In particular, IoT-devices need a new type of memory - faster and more energy-efficient and looks like IBM created fundamentally new one called PCM (Phase Change Memory). During phase transitions, memory has a high writing and reading speed and can withstand up to 10 million write cycles (that is several times more than conventional USB-devices.) This new memory can be used in high-performance systems that perform complex calculations. It will also be in high demand for situations when instantaneous data processing is required, for example, for IoT-devices to receive, process, and respond to a large amount of data from different sensors and smaller devices. This IBM memory also consumes very little energy.
The Data Analysis Market
IDC research firm presented its forecast of big data and business analytics market development. IDC pointed out that by 2019, the market will exceed $187 billion (for comparison, in 2015, its volume was $122 billion). In other words, the market growth in the coming years will exceed 50%. More than half of the turnover will come from web services while $55 billion will be provided by software. Analysts expect the market to grow in all three segments - infrastructure, software, and services. The main consumers of the big data and business analytics will be manufacturing, banking and finance services, healthcare, and utilities. The largest market for the data analysis sector in 2019 with the volume of $98 billion is forecast to be the US, the second place will be taken by Western Europe.
Twitter's Heron Analytic Tool
Data analysis for business today is a necessary condition for its success. Real-time data analysis is another must-have, which is especially true for so-called instant services - messengers or technical support services. Twitter has granted access to all to its real-time analytics tool called Heron. The tool has been developed by Twitter as a replacement for the Apache Storm solution over two years ago and its purpose is to allow users not only to have the ability to analyze their data in real time but also use it as a convenient tool to setup all processes. Initially, Apache Storm, developed by BackType marketing research company and bought by Twitter in 2011, was used for this type of work. However, despite all the Storm’s advantages (among them the scalability and the fault tolerance), it was hard to work with the system and hard to get good results. Heron is much more convenient and functional. The new solution allows increasing the “efficiency” up to 2-5 times (mainly due to lower operating expenses and capital investments).
Smart City Data Exchange
Data analysis is a type of technology, the scope of which is almost unlimited. Many large companies create entire departments and invest in data analysis development. For example, a few weeks ago, Hitachi launched the Hitachi Insight Group - a new direction of its business-oriented Internet solutions in the field of “Smart City.” More than $10 billion has been invested in the project. According to the company, industrial and manufacturing applications for the IoT brought the company an income of $5.4 billion and the company has 33 projects of this type going on. One of the first projects launched was the City Data Exchange - a new data market for “smart cities.” At the core of this tool lies the idea of combining the databases of public and private data providers for use in the smart city applications. The first city in which the project is fully implemented is Copenhagen. According to Hitachi managers, 100 suppliers of anonymized data (both public and private) have been connected to the system. The company will accumulate these data and provide access to them through RESTful JSON (API) as well as through a subscription service.
Heat Maps for VR E-commerce
E-commerce cannot sustainably exist without analytics instruments. With their help, the business owners can understand their clients’ actions, forecast the demand during a certain period of time or certain products, etc. One of such instruments is a clicks’ heat map which helps to watch what visitors of e-commerce websites pay attention in the first place as well as to identify blind spots and the places of high interest on their website. If you would like to create a free version of the heat map, you can use Yandex.Metrika or Mouseflow analytical services. Another player on the market - Littlstar - created a heat map for e-stores that use virtual reality. This map, as well as a standard heat map, will show where the visitors are looking. To get the data, Littlstar’s product analyzes the movement of the user’s head, using the statistical and cluster analysis. At the moment, such map is not available in the real-time regime and the developers explain that this is due to the fact that such heatmap requires more calculations and takes a few hours to complete.
The Future of Siri
Smart assistants are conquering our living rooms, smartphones, and browser windows. At WWDC conference, Apple is preparing to introduce a number of novelties. Among them, a toolset for Siri SDK developers, thanks to which it will be possible to use Siri voice assistant in external applications. In addition, it was reported that Apple is working on a device for smart home and voice assistant - an analog of Amazon Echo and presented recently Google Home. Apple’s assistant will be able to control the smart home components using the HomeKit platform and also by interacting with bots. Siri has a big advantage over the competitors - the system can operate in several languages already, which means a high potential interest for the future Apple’s development.
Amazon AI Cloud Systems
Amazon began testing a new service - the company provides its clients access to the artificial intelligence cloud systems. Amazon Web Services users will be able to use a number of additional features, among them - pattern recognition system, speech recognition and analysis systems, etc. Interesting to note that access to such functions was previously offered by Google, Microsoft, and IBM customers. Google, in particular, offers developers to create and control neural networks that can be then be used to analyze the data and a variety of other tasks.
No More "Buy" Button on Twitter
Twitter’s experiment of introducing an e-commerce element into the functionality of the microblogging platform was a failure. In late 2014, Twitter has launched a “Buy” button on the site that users could use to purchase goods without leaving the platform. However, the development of this feature will not continue - the e-commerce experts have left the team and the company’s e-commerce department will be disbanded. Regarding the decision, Twitter team commented that the button and the functionality haven't been in high demand and people aren’t really focused on shopping on social media sites. Facebook is also actively experimenting with this feature and so far it didn’t go further than experiments.
