12 Things about IoT Analytics every Technology CEO Needs to KnowPedro Uria-RecioBlockedUnblockFollowFollowingMay 19Gadgets in a shop in the Lowyat mall in Kuala LumpurThere are 12 things every technology CEO needs to know about IoT Analytics and data science for IoT.
IoT Analytics represents a quantum leap in the analytics world.
Did you know them?1.
The IOT market in 2018 was estimated at USD 151B, driven primarily by the manufacturing and the automotive sectors, consumer electronics, and China2.
Internet of Things (IoT) will cover things, data, people, and processes to get operationalized, becoming the Internet of Everything (IoE)3.
While many wireless networks are used, 5G is gaining attention because it is optimized for IoT.
4.
The Cloud and the Edge are getting entangled to host the IoT environment, effectively becoming the Fog5.
Open frameworks and vendor solutions are necessary for seamless Cloud to Edge integration6.
Semantic modeling enables interoperability among different IoT networks and devices7.
Existing data management capabilities need upgrading for dealing with the scale and distribution requirements of IoT8.
New Generations of Intelligent Devices optimized for machine learning, deep learning, natural language processing9.
Although cyber-attacks are not noticeably derailing IoT, data models must integrate data governance, Trusted HW/OS, cybersecurity, and Distributed Ledger technologies10.
AI can process a wide range of IoT information, including video, still images, speech, network traffic activity, and sensor data11.
IoT Data Scientists are increasing use optimized cross-platforms model compilers and developer-ready smart objects such as Echo12.
As IoT analytics becomes mature, a wide range of social, ethical and legal issues ariseLet’s go one by one:1.
The IOT market in 2018 was estimated at USD 151B, driven primarily by the manufacturing and the automotive sectors, consumer electronics, and ChinaAccording to “IoT Analytics”, a research firm, the global market for IoT was expected to grow 37% from 2017 to $151B in 2018.
Moreover, global technology spending on IoT is expected to reach $1.
2T in 2022, attaining a CAGR of 13.
6% according to IDC.
Within industrials, the IoT market growth is driven primarily by the manufacturing and automotive industries, followed by transportation & logistics.
These heavier industries lead in IoT by connecting the products that they manufacture or by connecting machines into a more efficient value chain.
On the consumer side, Amazon and Google have reached critical mass in connected homes despite security and privacy concerns.
Consumers, particularly younger ones, use these devices to initiate shopping, to control entertainment, or to adjust the temperature and lights.
This market dominance has significant implications for IoT strategy as manufacturers and retailers position their products and services to integrate with connected homes.
Finally, Chinese IoT firms are winning locally and starting to gain ground globally.
Many Western companies are trying to capture a small piece of the sizeable Chinese IoT market.
According to Ericcson, the number of cellular IoT connections is expected to reach 3.
5B in 2023, of which 2.
2 are in China and other North Asian countries.
However, reliable Chinese companies have emerged to compete, and Western companies are finding it more challenging than expected because these Chinese IoT firms are globalizing and moving into the Belt & Road initiative countries:Baidu, Alibaba, and Tencent in cloud infrastructure as a serviceXiaomi in wearables and smartphonesAyla in connected home ventilation, air conditioning, and appliances.
2.
Internet of Things (IoT) will cover things, data, people, and processes to get operationalized, becoming the Internet of Everything (IoE)IoT is expanding into what we call The Internet of Everything (IoE).
IoT brings together things, data, people and processes to make networked connections more relevant, valuable and actionable, resulting in increased capabilities, richer experiences, and more significant economic opportunities.
The Pillars of The Internet of Everything (IoE) are 4:Things: physical objects connected to the Internet and each other for intelligent decision making; often called the Internet of Things (IoT).
Data: using data to develop better insights and to make better.
People: connecting people in more relevant and valuable ways.
Process: timely delivering the right insights to the right person or machine.
As companies digitalize products and operations, incorporating things, data, people and processes, entirely new ways of doing business in industries emerge.
3.
While many wireless networks are used, 5G is gaining attention because it is optimized for IoT.
When the IoT market was nascent, solution providers, particularly of telecom operators, focused mainly on connectivity.
However, the work involved in connecting things and servers was just the beginning.
An increasingly paramount objective is to process the data that comes from connected things.
IoT connectivity involves balancing competing requirements, such as endpoint cost, power consumption, bandwidth, latency, connection density, operating cost, quality of service, and range.
No single networking technology optimizes all these requirements:Wi-Fi connectivityBluetooth Low EnergyNear Field Communication (NFC)Zigbee or other mesh radio networksSRF and point-to-point radio linksUART or serial linesSPI or I2C wired busesLora5G, the impending generation of cellular networks, is likely to gain attention in the IoT market because it is designed to optimize power consumption, bandwidth, latency, connection density, operating cost, quality of service and range.
Having said that 5G has a relatively high opex for IoT asset owners.
The use of 5G in mobile devices will change the landscape of the IoT, giving telecom operators a big push.
The Cloud, the Edge and the Fog (best-wallpaper.
net)4.
The Cloud and the Edge are getting entangled to host the IoT environment, effectively becoming the FogA common misbelief is that data need to be in the Cloud or some similar central location to be analyzed.
Sometimes this is true but performing some analytics at the Edge, closer to the devices that generated the data, is becoming an option.
In many industrial sectors shifting some analytics intelligence to the Edge may be more cost-effective.
Autonomous vehicles face a different challenge; even with better data-transport technologies such as 5G, unreliable response times may make edge-based solutions more relevant.
Five motivating factors for using Edge Computing is triplePreserve privacy: Data obtained by IoT devices can contain sensitive information (like GPS data, streams from cameras, or microphones).
With Edge Computing, sensitive data is preprocessed on-site, and only data that is privacy-compliant is sent to the Cloud for further analysis.
Reduce latency: When immediate results are needed, Edge applications can run machine-learning algorithms directly on IoT devices, and only communicate with the Cloud outside the critical path, for example, to continuously train machine learning models.
Robust to connectivity issues: Designing algorithms to run partially directly on the Edge ensures that applications are not disrupted in case of limited network connectivity.
Reduce cost: The Edge can be beneficial to reduce costs coming from expensive connectivity technologies.
Increase scalability: With the explosive growth of IoT devices, scalability becomes essential.
IoT will progressively move to Fog Computing or Intelligent Mesh architectures for scalability purposes.
Fog Computing evokes the meteorological effect of fog as a layer between IoT sensors (the ground) and cloud computing (the clouds).
This architecture addresses the IoT scale problem by inserting a gateway between the sensors and the data center that gathers data.
The Fog performs initial processing such as filtering and correlation before sending processed data to the Cloud.
The Fog could analyze and correlate events across various sensors and identify vulnerabilities.
It could then mitigate them by ignoring compromised devices.
In an Edge or Fog architecture, devices can be of three types depending on their role:Edge Sensors and Actuators are special-purpose devices that do not have full-fledged processors or operating systems and are connected to Edge Devices or Gateways via low power radio technologies.
Edge Devices are general purpose devices that can run full-fledged operating systems.
Edge Devices are often battery-powered.
For example, machines running Linux, Android, or iOS can qualify as Edge Devices.
Edge Devices forward raw or preprocessed data to the Cloud, like storage services, machine learning or analytics services.
Edge Devices also receive commands from the Cloud, such as configurations or queries.
Edge Gateways have the same function as Edge Devices but typically have unconstrained power supply and much higher hardware and software specifications.
For example, when you connect your Apple Watch to your iPhone via Bluetooth.
Your iPhone is acting as an Edge Device and your Apple Watch as an Edge Sensor and Actuator.
A fundamental part of it is the robust and seamless integration between the Edge and the Cloud; between the physical world and the logical world; between the flexible computing power of the Cloud and the instantaneity of preprocessing in the Edge.
As a result, an Edge application is made of several modules, each running at different places in the hierarchy:In Edge Devices, or even in sensors, a simple, lightweight set of rules can be used to filter, preprocess, aggregate or score IoT data.
Moreover, in a feedback loop, IoT devices receive commands to be executed on the physical worldIn the Edge Gateway, a machine learning module might be deployed to score pre-aggregated data.
In the Cloud, a more complex analytics module might be used to analyze data coming from Edge Gateways and Devices; A dashboard module might be deployed in the Cloud to provide a global data view or a query interface;An Edge Computing application should describe how modules interact and communicate, by clearly defining data flows between components, henceforth also defining visibility restrictions, business, and privacy rules.
Artificial Intelligence and Robotics will take the Edge to a world of self-service and operational efficiencies.
Edge autonomy is already a reality in a few verticals across automotive, healthcare, retail, manufacturing, among others.
For example, connected cars cannot afford the cost of latency, for the sake of detailed analysis.
Thanks to robotics, the Edge is becoming smarter, more self-reliant, is starting to have access to local storage and machine learning models or sets of rules to score the data.
It also can trigger actions, potentially involving other robotic devices.
5.
Open frameworks and vendor solutions are necessary for seamless Cloud to Edge integrationOver the past years, enterprises have distributed more computational power, storage, applications, and workloads onto disparate proprietary IoT and Edge platforms.
Public cloud providers are starting to provide management consoles in hybridized on-premises, Edge and Cloud environments, which will use AI to automate most of the real-time monitoring, anomaly detection, root-cause analysis, predictive analytics.
Furthermore, edge-based appliances will give public clouds a foothold in hybrid on-premise deployments, with products such as Microsoft Azure Stack, IBM Cloud Private, Oracle Cloud At Customer or AWS Outposts.
Cloud-to-edge interoperability frameworks will begin to take place.
Many industry initiatives are building vendor-neutral, open-source, loosely-coupled frameworks for distributing microservices to the Edge: like for example, EdgeX Foundry, Industrial Internet Consortium, the OpenFog Consortium, the Automotive Edge Computing Consortium, the Eclipse Foundation, and the Cloud Native Computing Foundation.
Ontologies like the ones used in semantic modeling were introduced by the Greek Philosopher Parmenides (wikipedia.
com)6.
Semantic modeling enables interoperability among different IoT networks and devicesIoT endpoints will grow into the trillions, come in all sizes, perform every imaginable function, and use a near-endless variety of APIs and protocols.
IoT endpoints might be a single sensor or may encompass an entire globe-spanning cloud of endpoints.
The IoT functions may include a dynamically shifting assortment of sensors, actuators, middleware, and machine-learning algorithms.
The more fragmented and heterogeneous the IoT grows, the more critical it is that all nodes and applications share a common data model.
Semantic technologies suggest a suitable approach for interoperability by sharing common vocabularies and enabling the interoperable representation of inferred data.
Semantic modeling uses ontologies to describe concepts and relationships between different entities that are connected using graphs.
An ontology encompasses a representation, formal naming, and definition of the categories, properties, and relations between the concepts, as well as definitions about individuals, classes, attributes, aspects, function terms, restrictions, rules, axioms or events.
In this standards framework, thing-level standards are like the clay from which pots of many shapes, sizes, designs, and uses may be molded.
Semantic models are not end-products.
They usually are only part of a solution and should be transparent to the end user.
The semantic annotation models should come with valid APIs and tools to process the semantics to extract actionable information from raw data.
Semantic interoperability standards will be essential.
As in any distributed environment, semantic interoperability in the Cloud will enable applications to understand the meaning of data element they import, acquire, retrieve, and receive from elsewhere, without confusing its meaning.
Another significant advantage of semantic modeling is that it makes data search extremely fast and scalable which is also relevant for IoT.
Several data semantic model proposals already exist.
Different standards development organizations have developed many of them.
Most of them are specific to some IoT vertical domains: like for example, the World Wide Web Consortium Thing Description, Smart Appliances REFerence, the Open Geospatial Consortium, the Open Connectivity Foundation, Schema.
org, the Constrained RESTufl Environments (CoRE) IETF working group, the IETF Thing-to-Thing Research Group (T2TRG) or IPSO Smart Objects.
7.
Existing data management capabilities need upgrading for dealing with the scale and distribution requirements of IoTAccording to Gartner, through 2019, one-third of IoT solutions will be abandoned before deployment due to a lack of data management capabilities adapted for IoT.
IoT solutions and the data they generate represent a significant shift in the requirements for storing and managing IoT data.
IoT is driving a substantial shift toward Hadoop and nonrelational forms of data persistence, that enable high-speed and high-volume data and event stream ingestion and distributed storage, with greater flexibility and cost efficiency.
Moreover, automated data management leverages machine learning capabilities and AI engines to make enterprise information management categories self-configuring and self-tuning.
These categories include data quality, metadata management, master data management, and data integration.
It automates many manual tasks and allows less technically skilled users to use data.
Chief Data Analytics Officers should evaluate the suitability of existing data management capabilities for dealing with the scale and distribution requirements of IoT deployment and the unique governance issues of IoT data.
In its most simple form, there are three key considerations:Firstly, speed would generally depend on how fast you need results:Hours: use a MapReduce technology, such as Hadoop or Spark for processing.
Seconds: send data into a stream processing system (like Apache Storm or Apache Samza), an in-memory computing system (like VoltDB or Sap Hana), or an interactive query system (like Apache Drill) for processingMilliseconds: send data to a complex event processing system where records are processed one by one.
Next, decide how much data to keep is a tradeoff between cost or risks and potential value of data:Keep and save all the dataProcess all the data in streaming keeping no data at allKeep a summarized version of the data, knowing that you probably cannot recover the original information againThe third question is where to do the processing and how much of that logic you should push towards the sensors:Do all processing at analytics servers or CloudPush some queries into the gateway or devicePush some queries down to sensors8.
New Generations of Intelligent Devices optimized for machine learning, deep learning, natural language processingAs of today, there are effectively three classes of devices:The simplest devices run on embedded 8-bit System-On-Chip (SOC) controllers.
An excellent example of this is the open source hardware platform, Arduino.
The next level up is the systems based on Atheros and ARM chips that have a minimal 32-bit architecture and runs an embedded Linux platform or dedicated embedded operating systems.
Most capable IoT devices run on full 32-bit or 64-bit platforms.
These systems may run a full Linux OS or another proper Operating System, such as Android.
Mobile phones are in this category.
The sensor market will evolve continuously.
New sensors will enable a broader range of situations and events to be detected, current sensors will fall in price or support new applications, and new algorithms will emerge to deduce more information from current sensor technologies.
Traditional memory architectures are not suited to all the tasks that endpoints need to perform.
Deep neural networks are often limited by memory bandwidth or processing power.
Over the past several years, chip vendors have introduced a new generation of hardware architectures designed for machine learning, deep learning, natural language processing, or image processing:New generations of GPUNew persistent-memory technologies, which represent a new memory tier that can provide cost-effective mass memory for high-performance workloads.
Tensorcore processing unitsField programmable gate arraysApplication-specific integrated circuits.
Edge requirements are driving the introduction of AI accelerators that are optimized for greater autonomy in mobile, embedded, robotics, and IoT devices.
New systems-on-chip support complex workloads such as real-time sensor-driven video, audio, speech, motion or locomotion.
This new generation systems-on-chip will be configured out of the box with various algorithms to help Edge nodes autonomously sense environments and respond effectively.
These systems will also improve availability, boot times and clustering methods9.
Although cyber-attacks are not noticeably derailing IoT, data models must integrate data governance, Trusted HW/OS, cybersecurity, and Distributed Ledger technologiesA person wears a Guy Fawkes mask, a trademark and symbol for the online hacktivist group Anonymous, ABC NewsThe number and diversity of devices and their associated applications is so large and is growing so fast, that the foundation of many of our cybersecurity premises is challenged.
Security models need to change to integrate broad-based network big data collection and visibility through dynamically applied controls.
Around half of CIOs admit they have been attacked.
Among whom, one fourth experienced severe damage as a result.
Even companies that have been significantly damaged are not considerably halting their IoT activities.
In a nutshell, cyber-security is a big concern, but not a barrier to IoT adoption.
Companies doing IoT at scale are investing more in cybersecurity without withdrawing back from their IoT activities.
IoT will inevitably involve a high number of things with not only inadequate security mechanisms and poorly written protocol implementations.
These low margin devices will contain minimal features and use the cheapest hardware and software available.
Having trusted hardware and operating system in the devices is paramount.
However, organizations often do not have control over the source and nature of the software and hardware used in IOT initiatives.
As a result, an IoT governance framework that guarantees appropriate processes in the creation, storage, use, and deletion of IoT information is also becoming increasingly important.
IoT Governance ranges from simple technical tasks to more complex issues, such as:Auditing devicesUpdating firmwareDisconnecting a rogue or stolen deviceUpdating the software on a deviceUpdating security credentialsRemotely enabling or disabling specific hardware capabilitiesLocating a lost deviceWiping secure data from a stolen deviceRemotely re-configuring Wi-Fi, GPRS, or network parametersControlling the source and nature of the software and hardware on the deviceControlling the use of the information on the deviceIoT cyber-security will enable companies to maintain a high level of data privacy and protection while ensuring reliable service delivery.
IoT cyber-security requires three steps to do this:Visibility: a real-time, accurate picture of devices, data, and the relationships between them scales our ability to make sense of billions of devices and their related information and interactions.
This visibility requires true automation and artificial intelligence; humans only will not be able to scale up adequately.
Threat awareness: the ability to identify threats based on understanding normal and abnormal behavior, identify indicators of compromise and to make decisions requires overcoming complexity and fragmentation in our environments.
Action: once we identify a threat or unusual behavior we need to act.
Acting requires the right technologies, processes, and people working together.
Additionally, blockchain for IoT can transform the way business transactions (including data exchanges and commands) are conducted globally through a trustworthy environment to automate and encode business transactions while preserving enterprise level privacy and security for all parties.
Blockchain is also playing a significant part in making transactions more seamless and faster and creating cost efficiencies in the supply chain.
Blockchain and other distributed ledger technologies provide decentralized trust across a network of untrusted players.
Blockchain is still a few years away from becoming a dominant mature technology.
In the interim, technology end users will be forced to integrate with the blockchain technologies and standards dictated by their prevailing customers or networks.
Blockchain is not the only distributed ledger technology.
There are others like Tangle.
The difference between Blockchain and Tangle is that blockchain requires consensus along the whole chain and in Tangle local consensus is enough to consider a transaction accurate.
Finally, from an analytics point of view Blockchains are a data source, not a database, and will not replace existing data management technologies.
10.
AI can process a wide range of IoT information, including video, still images, speech, network traffic activity, and sensor dataIoT vendors initially focused on creating applications and business models with IoT.
Now that some of these applications are taking off, AI is starting to play in IoT ecosystems.
Below are some of the most promising AI techniques used in IoT:Time series: Most IoT data are collected via sensors over time.
Often most readings are autocorrelated, which means profoundly affected by an earlier reading.
However, most machine learning algorithms (like Random Forests or SVM) do not consider autocorrelation and would often do poorly with IoT data.
Time series analysis algorithms (like.
ARIMA) and Recurrent Neural Networks (RNN) work better with IoT data.
Anomaly detection: Many IoT use cases like predictive maintenance, health warnings, alarms, optimizations, depend on detecting anomalies.
Anomaly detection poses several challenges.
Firstly, most use cases would not have training data, and hence unsupervised techniques, such as clustering, should be used.
Secondly, even if training data is available, anomalies are very scarce with a few hundred anomalies among millions of regular data points.
Thirdly, false positives are very common, and algorithms typically cannot explain why a data point is anomalous.
Therefore, after detecting anomalies, they must be understood in context and scrutinized by humans.
Graph analytics: this analytic technique allows the exploration of relationships between entities of interest such as organizations, people and transactions.
The application of graph processing and graph databases accelerates data preparation and enables more complex and adaptive data science.
Graph data can efficiently model and query data with complex interrelationships across data silos at a faster speed.
Reinforcement learning: this machine learning technique learns from interacting with the environment trying to optimize an objective, unlike supervised or unsupervised learning which typically used (but not always) offline data.
Therefore, reinforcement learning has heretofore played a central role AI in gaming and robotics, but beyond this, it has had far lower adoption in business.
It is precisely the opportunity that IoT brings to experiment and interface with the environment that is making reinforcement learning rule at the Edge.
Natural Language Processing (NLP): By 2020, 50% of analytical queries will use voice or text.
Voice and text will drive broader IoT adoption, allowing analytics tools to be as simple as a conversation with a virtual assistant at the Edge.
Facial analysis and recognition: Video cameras can now use AI to capture and understand human sentiment using facial analysis, for example in retail.
Other applications are related to national security where persons of interest have been identified and tracked in public places using CCTV cameras.
Another example is deploying a camera inside the truck watching the driver’s actions and facial movements to detect fatigue on the driver’s face and alert them immediately or to understand how often the driver is distracted.
Robotics: the combination of robotics with IoT devices and machine learning will take us to the next level of automation and operational efficiency.
Today’s retail stores or logistic centers are already employing robots for janitorial tasks, for taking stock of products from the shelves, or for back-office operations such as merchandising, goods movement and packaging.
Robots are also serving food in some restaurants.
Explainable AI: Most of these advanced AI models are complex black boxes that cannot explain why they reached a specific recommendation.
In some scenarios, businesses must justify how these models arrive at their recommendations.
Explainable AI provides an explanation of models in terms of accuracy, attributes, model statistics, and features.
11.
IoT Data Scientists are increasing use optimized cross-platforms model compilers and developer-ready smart objects such as EchoDevelopers of AI applications for Edge are doing their work in a growing range of frameworks and deploying their models to a myriad of hardware, software, and cloud environments.
This diversity complicates the task of making sure that each new AI model is optimized for its target platform.
Open-source AI-model compilers are now in the market to ensure that the toolchain automatically optimizes AI models for fast, efficient edge execution without compromising model accuracies.
Some of these compilers are AWS NNVM Compiler, Intel Ngraph, Google XLA, and NVIDIA TensorRT 3.
Moreover, developer-ready smart objects such as AWS’ DeepRacer, DeepLens, and the Echo family represent a paradigm shift in AI development for the Edge.
Going forward, more AI-infused edge applications, including robotics for consumer and business use, will be developed on workbenches that sprawl across both physical platforms such as these devices as well as virtual workspaces in the Cloud.
As this trend intensifies, more data scientists will begin to litter their physical workspace with a menagerie of AI-infused devices for demonstration, prototyping, and production development purposes.
Finally, AI developers will increasingly use microservices (APIs) for edge development.
Microservices enable pushing business rules, processes or policies and deploying new services quickly with other platforms.
Microservices can run at the Edge and support IoT platforms.
Edge devices will need to excel at delivery to give businesses performance.
12.
As IoT analytics becomes mature, a wide range of social, ethical and legal issues ariseIoT asset owners are increasingly restricting who can view and use data coming from their machines.
The company owning the data-producing asset, for example, may not be the company best positioned to leverage the data.
Maybe the Original Device Manufacturer (OEM) or the users are better positioned.
Disputes and legal wrangling over data ownership and access can delay value creation.
Moreover, many governments have implemented strict data sovereignty and privacy regulations, often for good reasons, but in practice are creating further restrictions and complications.
We are also seeing the emergence of IoT data marketplaces.
For example, insurance companies have tapped into connected cars for getting the history of its subscribers.
According to the Gartner IoT survey, 35 percent of consumers would agree to sell data collected by their products and services.
The need for building more sophisticated and reliable ML models will justify the need for more data.
So, naturally, enterprises will start purchasing trustworthy datasets through intermediate platforms that can offer reliable marketplaces.
IoT is transforming all business sectors, from consumer devices to large-scale manufacturing.
Therefore, as IoT matures, a wide range of social, legal and ethical issues will grow in importance.
An IoT solution shall be not only technically effective but also socially acceptable to be successful.
These include ownership of data and the insights made from it, algorithmic bias, privacy, and compliance with regulations such as GDPR (General Data Protection Regulation).
Chief Data Analytics Officers must, therefore, educate themselves and their staff in this area, and consider forming groups, such as ethics councils, to review corporate strategy.
They should also consider having key algorithms and AI systems reviewed by external consultancies to identify potential bias.
About the authorPedro URIA RECIOPedro URIA RECIO is a senior marketing leader highly experienced in data analytics, AI, product marketing, and P&L management.
In Axiata Group, Pedro built from scratch a business transformation team that drove revenue-generating projects across all business units.
By expanding this unit to a regional organization of 170 analytics and marketing professionals over 11 business units, he eventually became the Group’s head of analytics with a focus on marketing and data monetization.
References“The Internet of Everything: How More Relevant and Valuable Connections Will Change the World”, Cisco, Dave Evans“A Reference Architecture For The Internet of Things”, WSO2, Paul Fremantle“IoT Analytics: Using Big Data to Architect IoT Solutions”, WSO2, Srinath Perera“Ten trends shaping the Internet of Things business landscape”, McKinsey, Eric Lamarre, Brett May“IoT-Lite: A Lightweight Semantic Model for the Internet of Things”, Maria Bermudez-Edo, Tarek Elsaleh, Payam Barnaghi, Kerry Taylor“Internet of Everything (IoE) Market 2018 Global Industry Analysis By Share, Key Company, Trends, Size, Emerging Technologies, Growth Factors, And Regional Forecast To 2023”, Marketwatch“2018 Roundup Of Internet Of Things Forecasts And Market Estimates”, Forbes, Louis Columbus“How IoT Impacts Data and Analytics”, Gartner, Christy Pettey“A guide to Edge IoT analytics”, IBM, Andrea Reale“Data and analytics trends to shape 2019”, FutureIoT,“Top 7 Big Data Analytics Trends for 2019”, IoT Insights, Kamalika Some“Gartner Identifies Top 10 Strategic IoT Technologies and Trends”, Gartner, Gloria Omale“Six Trends in IoT and Edge Computing to Track in 2019”, Hortonworks, Dinesh Chandrasekhar“Four Predictions For Edge Computing In 2019”, Forbes, Frank Cittadino“Emerging frameworks for cross-silo IoT data models”, Internet of Things Agenda, Chonggang Wang“Defining the Unified Data Model for the Internet of Things”, James Kobielus“Predictions for Edge Computing in 2019”, IT Knowldge Exchange, Tom Curtin“Federating Unfettered Analytics Across IoT’s Sprawl”, James Kobielus“The Internet of Everything (IoE), bbvaopenmind, Ahmed Banafa“What is the Internet of Everything (IoE)?”, Cisco“Security for the Internet of Everything: Turning the network Into a giant sensor”, Network World, Steve Martino“Creating the Internet of Everything”, IoT for All“Ten trends shaping the Internet of Things business landscape”, McKinsey Digital, Eric Lamarre, Brett May Share“An Introduction to WSO2 IoT Architecture”, WSO2, Geeth Munasinghe“Semantic interoperability key to realizing IoT value”, IoT Agenda, Chris Drake“How do semantic technologies enable the Internet of Things?”, Semantics.
cc.. More details