Industrial automation systems have historically relied on digital elements installed locally to the controlled equipment and machinery. As computational capabilities progressed, users took advantage of the possibilities, often adding more analytical features.
Eventually, many users would search for even more advanced ways to access and evaluate the data residing in control systems. Partly this was for convenient accessibility, but in many cases these users wanted to add Industrial Internet of Things (IIoT) data connectivity and analytical abilities.
One development in this area is using cloud computing as the best way to connect with one or many automation platforms. Cloud computing is gaining traction quickly because it eliminates many of the barriers encountered when taking an idea from concept to product. In the past, the projected cost of computing hardware and network hardware, and the associated configuration labor and continuous maintenance, doomed many projects before they even got off the ground.
Cloud computing provides an alternative where these resources can be deployed at a relatively low cost, and then economically scaled up as the concept matures.
Integrating cloud computing with industrial automation creates some immediate concerns, while at the same time providing opportunities. Users are rightly worried about maintaining secure connectivity and reliable data handling. These challenges must be met head-on to realize the scalability benefits and fleet-wide advantages of using cloud computing.
Classic industrial automation systems involved sensors and actuators wired to the inputs/outputs of programmable logic controllers (PLCs), which in turn were connected to human-machine interfaces (HMIs). This resulted in high performance and satisfactory security, at least from a local standpoint. However, gaining remote connectivity to these on-site resources, accessing the data, and processing it required significant and complicated computing and networking resources.
Newer PLCs are now available to remove this pain point by connecting directly to cloud services, with no intermediate processing required (Fig. 1).
Connecting industrial automation sources to the cloud seems like a great solution, and it is successfully used in many consumer and commercial applications. But operational organizations are concerned about keeping their data from falling into the wrong hands, and they need to protect themselves against digital intruders.
The major cloud computing providers—Microsoft Azure, Amazon Web Services (AWS) and Google—are acutely aware of the need for security, and therefore provide their end-users with a host of security features for authenticating devices and securely transmitting data. They offer software development kits (SDKs) so users can incorporate proper authentication of on-premises devices with the cloud infrastructure.
Most cloud services also support the MQTT protocol using transport layer security (TLS). MQTT is a publish-subscribe protocol suited for efficiently moving data, even over low bandwidth and intermittent connections, and TLS is an industry-standard networking cryptographic protocol. Authenticating with cloud infrastructure doesn’t have to be complex, as there are plenty of off-shelf solutions.
For example, some modern PLCs natively use secure MQTT communications, or can interact with web services securely by pushing/pulling data over the secure hypertext transfer protocol (HTTPS). The most advanced of these PLCs may be natively certified to work with a cloud service like Azure, requiring only a few parameters—URL, user/password credentials and client IT—to establish a session (Fig. 2).
Once a control system has been configured to securely share data with the cloud, the next concern for many automation designers is their ability to guarantee their users can access the data and analytical results at any time. Self-performing management of cloud computing is a task outside the expertise of many organizations, but using established platforms such as Azure, AWS and Google can be the best answer (Fig. 3).
End-users also have options for defining how and where data is stored and processed. Critical data can be replicated in different regions so that if there is a catastrophe, nothing is lost. Redundancy is critical to many operations, but for many on-premises applications the redundant servers are in the same physical location.
With cloud computing, the redundant servers can be located in different parts of the country, or the world. This geographic diversity greatly improves the chances of staying operational in the event of a power outage, natural disaster and other unexpected disruption of service.
Many users have struggled to justify IIoT projects of any size due to the large initial outlay required for PCs and/or servers. It is possible to use intermediate gateways to consolidate some data, but this adds another point of failure. Instead, modern PLCs with cloud computing services is a two-part solution providing results greater than the sum of the parts. Modern PLCs can be designed into new projects or added to existing installations.
Traditional PLCs acted as unstructured data sources, with limited preprocessing functions. Modern PLCs include additional capabilities to aggregate, calculate, and create structured data, and associate with context—also known as metadata—before sending it to the cloud.
This preprocessing reduces the cost of storing, transmitting, visualizing, generating notifications and performing cloud-based analytics, and it simplifies the usability and repeatability of a solution. The transition of information from the operational technology (OT) source to the information technology (IT) destination is thus streamlined.
Cloud Computing Services
Using modern PLCs in this way reduces the cost of cloud services, which are offered a la carte. These services are dependent upon the types of structured data sources available from modern PLCs. Traditional PLCs could provide bulk data, but this would drive up transmission costs and burden networks, and it would shift the preprocessing burden into the cloud, where service charges can stack up.
Cloud computing services are completely scalable for applications, servers and data storage. Users can start small and grow the resources as needed.
A Team Approach
Many facilities are accustomed to a divide between OT personnel who understand the field equipment and IT personnel who are more proficient in networking and cloud services. Using modern PLCs with cloud computing, without the need for complicated intermediate hardware and software, allows much more overlap and collaboration in these roles, and OT personnel will find the IT technologies more accessible.
For instance, if more field data points are identified, the OT team can simply add them into the source PLC and then access the points as needed with the cloud services. OT maintains ownership of the application, while gaining valuable insight on IT technologies.
While end-users will realize many benefits to implementing cloud projects, there are additional advantages for machine and equipment builders. These OEMs can provide new equipment with an all-in-one PLC for local machine control, along with direct connectivity to the cloud. Some PLCs include multiple Ethernet ports so OEMs can establish one network segment for local machine connections, and a second for cloud/internet connectivity.
PLCs used in this role should also provide a robust instruction set to simplify building contextualized data structures that can be consumed natively by cloud infrastructures. The PLC should have sufficient memory to accomplish local data logging, which can be used to store data if there is a network problem, and then forward it when the network is restored.
Using off-the-shelf products and services in this way allows OEMs to design, build and deploy advanced IIoT applications with minimum effort and risk. Some cloud-based software as a service (SaaS) options available are for:
- Energy monitoring.
- Machine performance and availability.
- Overall equipment effectiveness (OEE).
- Quality metrics.
- Remote monitoring and notifications.
- Predictive maintenance.
- Data historians and trends.
Getting started is as simple as registering on a cloud vendor website, and then entering the authentication parameters into a cloud-capable PLC. OEMs can use cloud computing and SaaS to quickly add IIoT capabilities to one machine, with only a modest amount of required training.
These IIoT features can be extended to every machine they sell, as they are produced. The OEM can then monitor and compare the performance of an entire fleet of machines, located anywhere in the world, to improve their product and offer value-added services to their clients.
By removing the burden of learning and experimenting with many underlying technologies, modern cloud-capable PLCs are enabling OEMs and end-users to focus on what they do best, while still taking advantage of the latest technologies.
Damon Purvis is the PLC product manager at AutomationDirect.com. He has more than 22 years of industrial automation experience. Previous roles have included designing and deploying automated solutions in a variety of industries and managing product development of manufacturing data management and business intelligence applications.