IIoT Gateway

Gateways to Delivering the Value of IIoT

Yesterday afternoon I happened to look at the top bar of my computer screen and noticed I had twelve search tabs open. It is easy to get lost in an information labyrinth when you are researching the IIoT—the Industrial Internet of Things. And I wasn’t just researching everything about it—this was not a one-hundred-thousand-foot survey. I was focused on a particular functional area of the IIoT: the Gateway.

Motivating this research was my belief that IIoT gateways, metal or plastic boxes with circuit boards (PCBAs) inside them, represent a very large revenue opportunity for electronics manufacturing services (EMS) providers. The hypothesis: if the value of the IIoT is even half of what the business research community believes it will be, and given that the gateway is an essential link in the IIoT value delivery chain, then a whole lot of gateways will be needed as IIoT implementations begin to proliferate.

Gateways convert the data emitted by sensors in the factory or the processing plant (data sent in dozens of different protocols) to standard IP-formatted data, then send this data over the network for processing. The sexy part of the IIoT is analyzing these terabytes of sensor readings—“Big Data”—for patterns, correlations, anomalies, and other types of actionable information. But the plumbing needs to be there too, and gateways are just that—infrastructure without which there would be no digits to analyze.

So why was I getting literally amazed by the IIoT, with one web page leading me to another until I forgot what I was trying to do in the first place? If gateways are plumbing, and everybody needs plumbing, isn’t that enough to know?

My thinking was: it is not. I was trying to validate a secondary hypothesis: that the IIoT gateway will evolve from being mostly a translator to being an integral part of the IIoT’s computational fabric: not just a translator, but an author. This is important to know, because if true it means that there will be a continuous stream of new gateway designs needing to be built; new designs mean new PCBAs, which mean new business for EMS providers.

I never found the path out of the labyrinth, but I went far enough through it to be confident to state that gateways will inevitably become more powerful, and the number of functions they perform will increase. Innovation in gateway technology will drive many new design projects. Consider the following:

  • The filtering of sensor data will become more important. It is expensive to transmit and store data, so why send useless data to the cloud, the data center, or a “data lake” for analysis? Gateways will need to perform increasingly sophisticated filtering operations on sensor data, enabled by faster processors and more memory.
  • Gateways will be designed for multi-device data aggregation and real time analytics, and used in process control applications where the results of the analyses are needed quickly and used locally. An example of such an application is if the temperature of a holding tank in an in-line pharmaceutical manufacturing process needs to be changed based on an optimization algorithm running on the gateway. This is the Edge Computing model, or as some call it, Fog computing.(How about “Little Data”?)
  • Gateways will be expected to service an increasing number of sensors and smart devices (per gateway).
  • Functions associated with device discovery and provisioning will be delegated to the gateway. Device management will be shared with the IIoT management platform running in the cloud.
  • Gateways will be designed to network with each other and share resources. Driving this design improvement will be the benefits of increasing the reliability of the network via some degree of redundancy, and achieving the required amount of computing power from a distributed architecture.
  • The demand for increased security via more powerful encryption technology will cause use of larger and faster processors.
  • Gateways will become servers for applications that extend the value of real time analytics.For example, consider an incident management application that initiates a “blast” voice conference call connecting a pre-determined DL of subject matter experts best qualified to quickly respond to and manage the incident. (REDCOM offers such an application.)

The IIoT is a complex adaptive system, which means that gateways will both influence and be influenced by developments in other parts of the system.Advances in protocol standardization, smart sensors, software applications, and network technologies will affect gateway design.As a distinct computing device category, IIoT gateways are still in the early stages of their evolution, but even today, it is clear that they are destined to bear an increasing amount of the burden associated with realizing the value of IIoT.

IIoT Fog

Implementing IoT: Are You on Edge, or in the Fog?

Just when I started to think I had a basic understanding of the Cloud, a new type of weather has come rolling in: the Fog. I was dismayed when I first encountered it—another computing paradigm to learn about, sigh. But as it turns out, if you know what the Cloud is, it’s a lot easier to “demistify” this new term in the Computing lexicon. And if you are an Electronics Manufacturing Services provider like REDCOM EMS, you may even get excited about it—explanation below.

I didn’t go looking for the Fog—it found me. I wanted to know more about the Internet of Things, because everybody is talking about IoT, and I hate it when I can’t chime in to a conversation with a clever comment. As most readers know, you can’t have IoT without the Cloud, because the Cloud is where the data that comes from all the “Things” is analyzed and converted to useful information. This is especially true for Industrial IoT (IIoT) where there may be thousands of sensors and actuators networked to hundreds of machines in a large factory. These sensors can generate terabytes of data per hour.

So you can see that a big problem arises in IIoT: there is too much data. Sending all this “big data” straight to the Cloud is expensive and even with todays’ high-bandwidth networks, too slow to enable fast-enough communication back to the factory, (for example, to shut down an overheated motor). The reliability of the connection to the Cloud is also a concern.

The obvious solution is to pre-process this data, then upload just a sampled-down or otherwise permuted subset of it to the Cloud. This pre-processing occurs on the factory floor, and is accomplished by IoT gateways, sensor hubs, and “intelligent endpoints” designed and programmed for specific tasks. While a gateway is so named because it translates machine communication protocols to internet-related protocols, gateways can also perform other functions, such as analyses of smaller data sets for immediate, locally-useful interpretation and action.

All this seems pretty clear, so you might wonder how Fog gets into it. These gateways, located near or at the edge of the network, are the “fog” layer between the sky (the Cloud) and the ground (the sensors and nodes hooked up to the machines and other processing equipment). “Fog” is just an extension of the cloud metaphor, the term being closely connected to implementing IoT. “Edge computing” is a synonymous term, and some prefer it for its clarity compared to “Fog”. That said, I think “Fog” will beat out “Edge” computing, if for no other reason than “Fog” having its’ own consortium.

The Fog is where the hardware is in the IoT. And there will be a lot of hardware. The opportunity for designers to innovate in Fog hardware is gigantic, and this innovation will manifest itself in many new Printed Circuit Board Assemblies (PCBAs) mounted in gateway chassis. Common to these new designs will be multiple I/O connectors, MCUs (microcontrollers) and full processors, and devices that do not lend themselves to highly automated assembly techniques. As more processing power comes onboard, thermal management will be of increasing importance.

And this is why I’m excited about the Fog: REDCOM EMS, having a history as an OEM of high-interoperability telecommunications hardware products, is on very familiar terms with the techniques and challenges related to manufacturing these kinds of PCBAs and chassis. Real estate is usually in short supply, and that’s why two-sided SMT is so important; we have Selective Soldering equipment, enabling us to efficiently produce these mixed-technology (SMT + through-hole) boards. We’ve designed and built boxes with extended operating temperature ranges, learning much about how to “keep cool”, be it with careful heat sink selection or simply making sure that connecting cables do not interfere with convection flows.

REDCOM EMS is very well-positioned to service the needs of customers who, like us, are pretty stoked about IoT/IIoT, and believe that the Fog is clearly a huge opportunity. Together we can deliver on the promise of IoT/IIoT.


What Do Bruce Springsteen and IoT Have in Common?

I’m old enough to remember when the term “hype” started to work its way into the vernacular. It was 1973, and a guy named Bruce Springsteen cut an album called Greetings from Asbury Park. I ignored it–I was “into” Pink Floyd, The Who, David Bowie, Led Zeppelin, and Yes. What I could not ignore was how much interest there was in Bruce Springsteen: in print; on the radio, everywhere and often. Was the music that good? Wasn’t this just his record label (Columbia) salting the media mine, blowing the marketing budget on a risky bet? I wasn’t the only one who noticed this inordinate amount of promotion, and before long not-so-complementary things were being said about Springsteen and Columbia: “It’s pure hype”; “They’re just hyping him up”; “It’s just a hype job”. I completely agreed.

Jump to August, 1975. Springsteen releases Born to Run. I listened to it, and it wasn’t just good. It was great—Triple Platinum grade great. Yes, the artist was hyped, but there was no doubt that the hype was justified. And if a few die-hard skeptics were left, Bruce came out with Darkness on the Edge of Town and The River to settle the matter.

“Hype” still connotes negatives to me. I tend to think that real things shouldn’t need hype to be successful, or that there must be an element of sham, a lack of genuineness somewhere in the ecosystem of hype. That’s why when I first noticed all the buzz about “IoT” a few years ago, I was skeptical: who profits from this? Who’s pushing the concept? Amazon? Cisco? Qualcomm? This skepticism persisted until Wednesday 9/21/16. On this day, representing REDCOM EMS at the AmCon Design & Manufacturing Expo, I experienced the IoT equivalent of Born to Run.

Shortly after the exhibit hall opened, I had the opportunity to talk with the Director of Operations at an LED lighting solutions company. He was at the show to identify companies that could help him get his controller “online”, so that smartphones and other network endpoints could control the lighting. His customers were asking for it because they know it can be done, and they want the economy, convenience and the enhanced functionality of a networked solution. All communications will be via Internet Protocol (IP). “Wow”, I said. ”I think we are talking about IoT here, right”? “Absolutely” he replied.

Not long after this exchange, an enthusiastic young man approached our display table holding what I thought was the fob for his car’s remote keyless system. I asked what he was looking for, and he held up the fob and said “Somebody that can make the circuit board that goes inside this.”

He opened the fob to expose a small but very complex Printed Circuit Board Assembly (PCBA). He told me that the board contained sensors that detected the concentrations of potentially-harmful things encountered on construction sites: particulates, gases, high noise levels. The Use Case is simple: the construction company issues a fob to all workers on the site. The fob transmits the worker’s GPS and other data from the sensors to a local computer via a wireless connection. Also, each worker’s smartphone (issued by the employer or BYOD) is also connected to the network. Now imagine that one worker’s fob detects a too-high concentration of dust. Not only will his smartphone vibrate, but the network monitor will see this—which could be an early indicator of a dangerous cloud of dust starting to waft through the site. The path of the dust cloud may be inferred by the sensor/GPS data, and an alert can be sent to those workers in the path of the cloud.

This is, most definitely, IoT.

IoT has been hyped. But as was the case with Bruce Springsteen, just because it’s been hyped, doesn’t mean that it isn’t real, exciting, and profoundly important. I am now a believer, and fairly sure that IoT will go on to produce a long series of hits.