The Cloud Services that Made for the IoT

Without the cloud services, the value of massive data created by the IoT devices cannot be realized. The great cloud computing power has shed light on the understanding of the data and generated business insights, and this is what makes IoT thrive in this digital age. Now it’s a good time to get an overview of what IoT cloud platforms can offer and how choosing the right cloud platform is the key to a successful IoT strategy.

There are two basic models in connecting the devices to the Internet. One is the device can connect the internet directly. Another is a gateway device that communicates with all devices and acts as a centralized process center in which it does the data processing and translation before connecting to the Internet. Other time, devices may also communicate with each other and then connect to the Internet directly or indirectly.

There are many IoT Cloud platforms solutions out there, which include the big companies that offer end-to-end services, the open source platforms that offer low cost solutions, and the friendly platforms of DIY hardware aim for IoT hobbyists.

Let’s compare some of these IoT cloud platforms in features and advantages.

Cloud Platforms Provided by Big Companies:

Samsung ARTIK Cloud

The ARTIK offers built-in services that are easily connected between diverse devices and cloud services.  One of the advantages is that on each step of the way, the ARTIK Cloud services expose its features as APIs, and hide the complexity inherent in IoT development.  Like other end-to-end IoT solution, the platform unifies hardware, software, and partner ecosystems in one friendly ecosystem.

GE Predix     

The Predix Platform is designed toward for Industrial Internet, which provides a wide range of built-in industrial microservices that enable business to increase productivity and minimizes hardware management. The primary components of the Predix systems consist of Predix Machine, Predix Services, and the Predix Cloud. The Predix Machine serves as gateways which collect data from devices and pushes to the Predix cloud.  The advantages of GE Predix platform are that intelligence can be built-in at the Predix Machine level where an automated decision can be made and an action like shutting down a specific device can be performed.  This is good, especially it did not require the connection to the cloud to make a decision especially if the network is not available. With open source Cloud Foundry microservices architecture, the Predix platform also supports the continuous improvement of industrial business processes.

Oracle IoT Cloud Service                    

Oracle IoT Cloud Service provides a Java friendly environment, which means that your existing knowledge and skills can bring your devices to the cloud and minimizes the time and cost to adopt your IoT strategy since you can install Java application in Oracle IoT Gateways and use Oracle Java ME Embedded to run on ARM or x86 devices.

One of the advantages of Oracle IoT Cloud is that it simplifies the implementation of Edge Computing (Fog) significantly by using the Device Virtualization model or the Digital Twins.  The Digital Twins features that there is an exact copy of the physical object exists in the system that undergoes the exact process as the physical counterpart. Since it is intrinsically intelligent enough just like the real thing, it can generate appropriate alerts without the need to write a separate gateway application that processes this. It acts like the real thing. If this works properly, you can take it to another level with the Predictive Twin, where you can virtually manipulate the timeline and see into the future and predict possible outcome in the alternate future.

Amazon AWS IoT        

The Amazon’s AWS IoT offers a complete IoT solution from hobbyists to industrial users. It has its own real-time operating system, Amazon FreeRTOS, for compatible IoT devices, such as the small, low-power microcontrollers (MCU). This FreeRTOS is derived from open-source RTOS but added more libraries and security codes specifically designed to enable the devices to have many strong features. It allows a more secure connection with encryption capability and supports Over-the-Air (OTA) firmware update. As a result, the AWS IoT enables secure, bi-directional communication between Internet-connected things and the AWS Cloud over a variety of protocols, such as HTTP and MQTT.

For middleware, AWS Greengrass is software that let you turn other local devices into powerful gateways.  Even when not connected to the Internet, it keeps data in sync.  It also enables devices to use AWS Lambda functions and extend the cloud capabilities to the local devices. Best of all, it allows devices to access the management capabilities of AWS IoT Core, which is the third IoT stack of Amazon’s IoT solution. 

The AWS IoT Core integrates the needs to connect all devices from edge and gateway devices to other powerful backend cloud services like AWS Lambda, DynamoDB, and machine learning.  It is powerful enough to support billions of devices and trillions of messages.  Amazon’s AWS IoT solution does provide an end-to-end IoT strategy for businesses at all levels.

Microsoft Azure IoT       

Like Amazon’s IoT solution, Microsoft is also a leader in providing a complete IoT solution for the industry.  Similarly, Microsoft offers its own real-time operating system to microcontrollers (MCU) as known as the Azure Sphere.  Although it’s Linux-based operating system, it has many built-in capabilities that embedded on the chip, which includes network connectivity and security subsystem called Pluton Security Subsystem. Microsoft set a high bar for its security solution from the ground up.  Unlike the Amazon’s FreeRTOS, Microsoft’s design at the hardware level is a new crossover class of MCUs that are made Linux-ready while acting like a SoC (System on Chip) provides application processor capabilities. It’s a hybrid of MCU and SoC.  Regardless of whether or not the device is this new type of MCUs, Azure IoT Edge can be deployed on both Linux and Windows for ARM and X64 architecture from an industrial gateway class device to Raspberry Pi or equivalent device. In addition, Azure IoT Edge allows Microsoft Cognitive Services to the local devices.  It truly brings the AI to the edge devices.

Azure IoT Hub is a middleware where it can manage a huge group of edge devices and supports many protocols, including HTTP, MQTT, AMPQ, or the Azure IoT Protocol Gateway for custom protocol. Like Amazon’s IoT, Azure IoT Hub can manage billions of IoT devices and messages.  In addition, Azure IoT Central provides a fully customizable IoT management without coding or developers to support it. The UI allows users to connect and set IoT devices’ properties and send telemetry and many more features.  

For industrial users, Azure IoT Solutions Accelerators can help them do things like remote monitoring, predictive maintenance and test thousands of simulated devices, especially at the development phase. Instead of building a custom IoT solution from scratch, it provides you with ready-to-deploy templates and common IoT scenarios application code.

Like Oracle, Microsoft has Azure Digital Twins platform that enables a virtual representation of a physical device lives in a virtual environment.  This powerful concept of the “digital twin” allows IoT projects to design, build, and view all aspects of development and maintenance phase.  It truly brings IoT development to another level especially it is integrated with other Microsoft services like AI and Mixed Reality.  For example, you can test your concept of IoT solution in different environments like factories, hospital, and offices and visualize the IoT solution in the virtual world and assess its efforts and issues. Microsoft’s IoT solution not only encompasses the current need of a complete end-to-end IoT implementation but also empowers businesses the abilities to explore other business opportunities that previously unrealized in the IoT arena.

Google Cloud IoT        

Similar to Amazon and Microsoft, Google has its special sauce that builds from the chip level and the Edge TPU, Google’s purpose-built ASIC, allows an end-to-end, cloud-to-edge IoT architecture.  It allows edge devices to have machine learning capabilities and deliver deep insights faster.  Like AWS Lambda or Azure Function, the Google’s Cloud Functions allow IoT devices to run or invoke other services without a need of a server. The Google’s IoT solution is like Azure, where it has an Edge IoT Core, Cloud IoT Core. But inside the Cloud IoT platform, it has the Cloud Pub/Sub program to support large scale message ingestion. Once the data is captured by the IoT devices by its Cloud IoT Core, the results are published to Cloud Pub/Sub, and further data analysis can be done in the cloud using Google BigQuery and addition machine learning can be processed with Cloud Machine Learning Engine.  Finally, users can use Google Data Studio to visualize and customize your reports from the IoT data.

Within the Cloud IoT Core services, there are two main components: Device Manager and Protocol Bridge.  The Device Manager allows each device to have a secure identity, and to be monitored and managed securely. The Protocol Bridge makes secure communication possible in which it provides connection endpoints using MQTT and HTTP protocols and provides automatic load balancing.

The Google Cloud IoT solution can manage billions of sensors and edge devices.  It allows users to monitor devices and know their conditions with its real-time tracking capability. With AI built-in on the edge devices and in the cloud, predictive maintenance will be at ease since it can predict downtime and detect anomalies.  Actions can be triggered ahead of time.

Despite Google has a late start on offering its IoT services in comparing with Amazon and Microsoft, Google quickly leverages its existing cloud platform services to create a complete end-to-end IoT solution that is competitive in the IoT service market.

IBM Watson IoT       

IBM’s Watson IoT Platform is designed for enterprise IoT needs.  Like many good IoT service providers, Watson IoT Platform takes security to another level, it has a strict authentication mechanism that prevents impersonating devices. The platform allows Arm Mbed OS, an open-source embedded OS, devices to connect to IBM’s cloud via Arm Mbed Cloud Bridge, which simplifies processes of message translation, protocol mapping, and event mapping. Once connected to the IBM Cloud, it can access IBM Watson analytics services and cognitive services.

Like Amazon’s Lambda and Azure Functions, in Watson IoT Platform, it uses Apache OpenWhisk to launch serverless app and it is based on event-driven model. On the other hand, Bluemix’ Cloud Foundry, now owned by IBM, can be used to interact Bluemix Cloud services. Like GE’s Predix platform, you can use Cloud Foundry command line to deploy apps and services. Users can use both Cloud Foundry and OpenWhisk together as well since Cloud Foundry works great for longer-running actions or functionality, and OpenWhisk works on simple and quick functions.

IBM continues to develop its IoT platform, at the time of writing, Watson IoT Platform introduces Watson IoT Platform Edge in preview, which allows the abilities to create custom services to let edge devices to perform certain actions like failure notifications or making predictions. The IoT architecture will become more streamlined and comparable to other leading IoT providers, where it will consist of Edge Node, Edge device, Edge gateway, Edge services, Edge Services Catalog (Core IoT Service) and docker container images file repository called File Server Repository.


Headquarter in France, Atos IoT solution is driven by Industry 4.0, which focuses on expertise on manufacturer environment and settings.  With its rich partner ecosystem, Atos Cloud platform takes advantages of providing end-to-end professional consultation and strong partnerships with other cloud providers. Together, Atos offers businesses a practical transition path to custom their own Industry 4.0 operation and strategy.


Open Source Platforms

ThingSpeak Platform -

SiteWhere -

Macchina Platform -

Kaa Platform -


DIY Friendly Platform

Initial State Platform -

Carriots -

Losant -

Ubidots Platform -

Temboo -