Use of industrial microcontrollers: webcast questions answered
More answers are outlined below from the live audience Q&A in the June 8 webcast, “IIoT webcast series 2017: Sensing and control at the edge: microcontroller kits."
Control Engineering content manager Mark T. Hoske answers some questions about the use of industrial microcontroller kits that were not answered in the recent June 8 webcast, "IIoT webcast series 2017: Sensing and control at the edge: microcontroller kits," part of the IIoT webcast series 2017 from CFE Media publications Control Engineering, IIoT for Engineers, Oil & Gas Engineering, and Plant Engineering.
As explained in the webcast, microcontroller kits are emerging as an industrial solution for building Industrial Internet of Things (IIoT) digital devices and interactive objects, as the automation industry takes increasing note of these commercially available, edge-computer, board-level embedded controllers. Besides updating automation for new generations of software-savvy engineers, these low-cost elements make it possible to equip smaller systems with monitoring, control and optimization functions—just as SCADA enabled the automation of mid-sized systems decades ago.
What are the benefits, in terms of performance, of edge computing?
Hoske: Edge computing can put the decision part of the control loop (sense, decide, and actuate) closer to the process for greater, reliability, security, and flexibility, depending on design. Pre-processing of data also passes along a subset of data, closer to actual information, rather than all sensor data, decreasing the load on communications, to streamline and help optimize additional analysis and decisions.
Security is a big concern, and it appears you are saying that edge computing and open-source increase security?
Hoske: It depends. Newer architectures can be updated for security threats regularly and more easily and can fit into a design that considers cybersecurity concerns from inception, unlike many systems now in place. Frequent personnel training at every level is part of the process. As with safety, everyone is responsible.
Clearly, cybersecurity will need to be addressed in any open implementation. In the webcast, Arlen Nipper, co-founder and president /CTO of Cirrus Link Solutions, cited a 2017 Eclipse Foundation IoT developers’ survey where 20% of respondents deploys IoT solutions using an open hardware platform, and 31% use an open hardware platform for prototyping IoT solutions. In separate questions in the same survey, Linux was used as the operating system in 67% of IoT gateways, and HTTP and MQTT were used for more than 60% and nearly 55% of messaging protocols for IoT solutions.
Years ago, when fieldbus was introduced the PID controller could be found in the valve positioner. How is that different from the edge microcontroller?
Hoske: It’s a similar idea, as I mentioned when comparing edge controllers to DCS, though today’s processors are faster and networks more capable and flexible, with easier configuration and updates. While putting a controller in a device near the process might be a similar concept, hardware and software certainly differ. Nipper cited popular edge computing hardware (Raspberry Pi, BeagleBone, and Arduino) and software components (node.js, npm, MQTT.fx, Eclipse IoT, Mosquitto, and Paho) that may be used in newer edge computing, IoT-related applications.
Can you suggest an open comprehensive course on IIoT?
Hoske: Control Engineering and Plant Engineering have hundreds of articles on the topic, other webcasts, and research. Both sites also have an IIoT pull-down menu, upper left, which is dedicated to the subject. More is planned, also, at CFE Media’s education site CFE Edu.
Any thoughts on adding more intelligence at the edge to operate when network degrades or breaks? Humans can continue to operate when communications degrade. Why not more intelligence for machines at the edge? Even at subassembly level?
Hoske: Certainly, putting cost-effective redundancy on the edge can be strategy to add reliability for critical processes, and ensure actionable information is provided to those who need it to make better decisions more quickly. Working from wider goals or from points of constraint can often help ensure designs are appropriate to serve needs locally, at the site, organization, and throughout the supply chain. Applying technology because it’s cool may produce some good results and encourage others to update, but doing so for the right reasons is better and more sustainable.
Nipper noted that the $30 board-level microcontrollers are not ruggedized and don’t have redundancy, or embedded security. More rugged "industrial" versions are becoming available, he added.
See the archived version of this and other webcasts for one year after the airing date, at www.controleng.com/webcasts.
Edited by Kevin Parker, webcast moderator and senior contributing editor for CFE Media, email@example.com.