Cloud Computing vs. Field Controllers | Automation.com

Cloud Computing vs. Field Controllers

November 102014
Cloud Computing vs. Field Controllers

Bill Lydon’s Automation Perspective

By Bill Lydon, Editor

Some industrial automation vendors and computer technology providers are offering cloud services that include historians, analytics, and complete MES (Manufacturing Execution System). Certainly, these services offer a new way to improve industrial automation, control, and operations by applying more computing power. Another important trend is leveraging new, more powerful controllers within the secure automation system environment.

Cloud Historians & Analytics

Cloud computing leverages shared resources and economies of scale similar to an electric utility. The Cloud delivers powerful computing power and massive storage on demand. Companies only pay for what they need. This model transforms a traditional capital expenditure (CAPEX) of investing in dedicated hardware to an operating expenditure (OPEX) where you "pay as you go" or pay for use. One stated advantage is cloud computing enables users to focus on projects instead of infrastructure and software administration details. Cloud storage can be used as plant historians. A very interesting new development is analytics in the cloud which is offered by Google and Microsoft.

The Google Analytics Measurement Protocol allows developers to make HTTP requests to send raw data directly to Google Analytics servers, store it and use the rich set of Google analytics to analyze it. Knowledge of XML and programming is required. Lantronix provides an example of sending data from a digital scale via a Lantronix device to Google Analytics.

The Microsoft Azure Machine Learning is more refined and offers the ML Studio integrated, drag-and-drop development environment. Users can create analytics with a library of sample experiments and sophisticated algorithms from Microsoft Research, plus data flow graphs to define relationships. This is a powerful tool to mine data and create optimizations, predictions, and tune operations. Azure Machine Learning reminds me of Visual Basic that demystified programming and enabled a wide base of users to create solutions tailored to their needs. Potential users can try AzureML for free.

Cloud Manufacturing Execution System (MES)

Cloud-based MES simplifies deployment and lowers the cost of implementing a system that increases manufacturing efficiency. This enables smaller companies to achieve operating advantages and efficiencies that previously only large companies enjoyed. For example, Howard Hauser, VP Operations of Hiawatha Rubber, described how their small private company has orchestrated the use of modern manufacturing technology as part of their competitive strategy to improve the operations and pursue new markets. Their cloud-based system provides a range of functions, including product lifecycle management (PLM), enterprise resource planning (ERP), order entry and tracking, manufacturing execution systems (MES), and supply chain management (SCM). The cloud-based system from Plex Systems was implemented in just five months.

Cloud Constraints

Cloud computing does have some constraints and risks.

Communications

The most obvious constraint is the dependency on persistent high-speed communications so servers can interact with the manufacturing facility. This limits use to functions that can live with interruptions. Manufacturing automation processes that rely on tightly-coupled, high-availability systems cannot risk the loss of communications with cloud computing servers. Companies that store historic data and perform analysis for functions such as predictive maintenance and macro level process optimization can benefit from cloud computing. That being said, I am always amazed at the speed in which Google returns answers. For example, I did a search for “enthalpy” and it returned results in 280 milliseconds.

Cyber Security & IP Risk

Cloud vendors are working hard to protect data, but sending information to outside sources has inherent risks. Compounding this is the risk of a company’s intellectual property being stolen by competitors or imitators. This is a very difficult issue and cloud suppliers are adamant that they provide safeguards to protect users’ information. When I ask providers if they will insure the user for damages due to lost data, the answer is an emphatic, “NO!” The user ultimately has to make a risk assessment based on a number of factors.

Driving Computing to the Edge

Powerful industrial controllers now incorporate multicore processors. These controllers can perform analytics and collect short term historian data close to the process within the protected plant automation system environment. This new breed of controllers support multiple communications ports, protocols, wireless, embedded web servers, embedded historians, analytic engines, email servers, and web services. The Internet of Things (IoT) trend of driving communications to the edge (sensors, actuators, etc.) also embraces more computing power at the edge devices. Hadoop technology is making big data applications practical for collecting and analyzing huge amounts of data. Hadoop leverages a large number of machines that don’t share any memory or disks. The latest powerful industrial controllers are platforms that can support these technologies.

New Models

The Internet of Things (IoT) and Industry 4.0 concepts embrace a number of newer technologies, including cloud computing, powerful automation controllers, and Hadoop. These concepts are starting to change automation system architecture by leveraging COTS (commercial off-the-shelf) technology.

Related Articles & Information

MORE ARTICLES

VIEW ALL

RELATED