When it comes to creating a digital enterprise and connecting your business to various cloud-based applications, it’s important to understand the various platforms that make it all happen. For the Internet of Things (IoT), the ARC Advisory Group recently identified four new platforms that are needed for your digital enterprise: The Device Connectivity Platform, Cloud Computing Platform, Cloud Application Platform, and the Cloud Analytics Platform.

The Device Connectivity Platform

This first platform consists of the hardware and software needed to actually enable connectivity between the cloud and your field data. Devices like Moxa’s IIoT Edge Gateways perform this function, while also providing you with data acquisition from OT devices, data filtering or pre-processing, remote device management, an additional layer of security, and more. Additionally, built-in API support can provide easy and seamless connection to a range of cloud-based applications.

The Cloud Computing and Cloud Application Platforms

After the Device Connectivity Platform, the next two platforms can be summarized as Infrastructure as a Service (IaaS). Major players like AWS, Microsoft Azure, Google Cloud, and Alibaba Cloud are making large investments in their physical computing and cloud storage business, specifically in the industrial space. More importantly, they also provide the software platform that a wide range of applications can be built upon. When reviewing options for your business, it’s important to consider compatibility between your IIoT Edge Gateway and your selected cloud application.

The Cloud Analytics Platform

The final platform emerging for the industrial space revolves around analytics. With more and more machines being connected each day, the amount of associated data has grown rapidly. This has led to increased demand for effective data analytics, machine learning, and even artificial intelligence. There is a lot of competition in this space, including options such as IBM Watson, Amazon AI, Microsoft Cortana, and Google DeepMind, to name a few. Selecting the right application will be key to successfully harnessing your data and turning them into real insights for your business.

So what does this all look like from end-to-end? In recent years, our team at Moxa has seen some similar architectures emerging from these applications. Continue on to our next article to find out what some of our customers have done in the past.

Emerging IIoT Architectures

When implementing various IoT technologies into your business, data acquired from your operations is generally transmitted and stored on third party hosted cloud servers. This approach is very different from the traditional enterprise IT approach, where information is saved on local servers. By leveraging third party cloud services, OT device data and overall system health can be remotely accessed, monitored, and managed by connecting to the cloud server through the Internet. However, as more businesses embark to adopt Industrial IoT technology, they quickly begin to realize that there are many different approaches to connecting their edge data.

In this article, we highlight three primary IIoT architectures we see emerging with our customers, as well as how they work and the pros / cons we’ve identified for each.

Edge to Cloud

Edge to cloud is a common architecture that utilizes IIoT Edge Gateways to act as a middleman between edge devices and the Cloud. The IIoT Edge Gateway performs several important tasks, including translating industrial protocols, processing data locally, and filtering and transmitting selected data to the Cloud (which can be public or private).

Several industries are adopting this architecture, as it is similar to the traditional M2M or remote monitoring applications. However, with this approach, customers are adding IIoT Edge Gateways to the existing infrastructure and relying on Cloud infrastructure to provide device management, storage and dashboards for visualization. Many also plan or are already utilizing the data analytics and even some machine learning capabilities that these platforms offer. These features are great for gathering insights to your connected devices and operations performance, but also come at an added cost. The cost will typically vary depending on the amount of data that needs to be processed and where your operations are located.

Pros:

• Highly scalable
• Centralized storage and data visualization
• Powerful analytics through Cloud platforms

Cons:

• Quality of data accuracy (large applications with Big Data processing)
• Potential latency, affecting real-time data
• High bandwidth costs as the application grows

Additional costs to access advanced analytics modules from the Cloud platform

Edge to Modern Web SCADA

The next popular architecture is a new trend of hosting SCADA solutions on the cloud instead of local computers, often using a Software as a Service (SaaS) business model. These online Web SCADA systems would likely be hosted on a popular cloud service such as Azure, AWS, or Google Cloud Platform, speak today’s Restful Web APIs, and often use the same MQTT or AMQP protocols we see in Edge to Cloud architectures.

Although most industrial customers are still using the traditional model of in-house SCADA systems, we are witnessing sections of operations switching to Web SCADA systems so they can take advantage of the benefits provided by this new architecture. To help you understand whether the “Edge to Modern SCADA” model is right for your operations, here are some pros and cons to consider.

Pros:

• Significantly lower implementation and maintenance costs compared to physical SCADA systems
• Very fast deployment
• Easy access to real-time and historical information for multiple parties
• 3^rd party technical support
• Scalable server space with a pay-as-needed model
• Automatic updates to SCADA software

Cons:

• Increased cyber security threats
• Bandwidth limited to your ISP service

Dependent on the reliability of your cloud service (Azure, AWS, etc.)

Edge to Traditional SCADA

The third common architecture we are seeing is Edge to Traditional SCADA. One important area for this type of architecture are the APIs. Traditional SCADA systems must have their Polling protocols converted to new RESTFUL web APIs, or even better to Publish/Subscribe protocols with State fullness such as MQTT.

In this third architecture, we are seeing large industrial providers including GE, Schneider and Siemens offering their own Cloud software solutions.  Many of these are hosted on AWS or Azure, but those details are often abstracted from the user. In other cases, we are seeing some of these customers connect directly to Azure or AWS.

Pros:

• Full control of SCADA system and network
• Increased security, especially from cyber attacks
• Increased reliability and system availability

 Cons:

• Significant upfront investment
• Maintenance is tedious, time-consuming, and requires access to SCADA programmers
• More difficult and costly to expand upon

It’s important to note that the 3 architectures shown above can be much more complex in reality. In many cases, customers might use two or even three of these architectures in deployments, or they have various combinations of Cloud and SCADA solutions.

To enable any of these architectures, you need IIoT Edge Gateways that can handle selected data acquisition locally, while connecting your edge devices to cloud or web-based systems.