Industrial Internet of Things: Explained



Is this the start of a new productivity age?

A lot of us are tired of hearing about the Internet of Things (IoT), and the Industrial Internet of Things (IIoT) can also get swept under the rug with it. Perhaps this is because our experience with IoT has been wearable fitness devices (on the decline), wireless TVs, and tweet-activated candy dispensers. We may see further IoT advances that do more than make our lives slightly more convenient, but IIoT takes it to a whole other level, squeezing out the last drop of productivity, efficiency, reliability, and quality results in places like warehouses, manufacturing plants, energy generation plants, and much more. IIoT is what a lot of companies are gunning for.

The Industrial Internet of Things (IIoT) is nascent, and according to Intel®, General Electric (GE), Cisco, and others, we are on the cusp of a new age in productivity via copious intelligent sensors that constantly monitor a collection of sensor data, such as magnetometers, flow meters, pressure transducers, photodetectors, chemical sensors, camera images, and any other sensor that can supply an output that can be transmitted wirelessly. Let me explain in detail what Intel and GE are doing to build jet engines and run a boiler plant more efficiently. First, however, why is this happening now? We’ve had the Internet for a long time.

Figure 1: The Forecast for Connected Devices.

Figure 1: The Forecast for Connected Devices.

How Did We Get Here?

The tipping point for IIoT was clearly in advance of computing at greater power efficiencies, which has demonstrated Moore’s Law. As integrated chips get smaller and use lower supply voltages, power efficiency naturally increases. What has triggered the IoT wave is the recent combination of:

  • Greater computing in smaller packages; something of a tipping point for IoT.
  • Improved power efficiency, which is getting more difficult to improve upon.
  • Lower cost of sophisticated electronics.
  • Lower cost to implement via wireless communication rather than cabling.
  • Improved and lower cost solar or other energy harvesting methods.

All of the above have enabled the industrial sector to push some responsibility for computing and communication down to “the edge,” or physically inside the IIoT sensor. Large-scale IIoT has to be meticulously planned first, with people who know: what collection of data should trigger headend operators, the alarm levels that when combined with information from other sensors, give the collection of data meaning (e.g., what happens well before an impeller is cracked), what a full learning set should look like for matching patterns, and how this is all arranged in the data base in the cloud.

The sensor itself uses algorithms to look for specific patterns (pattern matching), might be able to perform object recognition, understand what specific variations in pressure, for instance, are too low in an anesthesia machine, complex combinations of dynamic data, and so forth. Facilities operators will groan, because as sensors get more complicated, they demonstrate bugs and are harder to replace. And this is going to be a growing pain of implementing IIoT. It’s possible that the entire set of plant-resident IIoT transmit to an IIoT service company far away, but setting up will require the tribal knowledge of the local staff.

Figure 2: GE’s Predix IIoT cloud platform connects machines, implements services, and delivers and scales apps. (Sources: www.predix.io)

Figure 2: GE’s Predix IIoT cloud platform connects machines, implements services, and delivers and scales apps. (Sources: www.predix.io)

There might be intermediate or local clouds, which are also referred to as “edge servers” and “fog servers.” An edge cloud is located in between two other networks, such as an edge cloud that is situated between the Internet and a private network. Fog resides in local to IIoT devices and can be used to enable real-time virtualization.

An IIoT “winning combination” or other less urgent items might be decided locally in the plant by a server. But the IIoT cloud requires additional, detailed engineering: software, integration with hardware, data organization and more. A larger platform then assembles the data with even more algorithms to create a cohesive picture that reveals the type of information that humans simply cannot create on such a grand scale.

Figure 3: The Intel® Quark™ SE is the processor for the Intel® Curie™ Module, meant for IIoT.

Figure 3: The Intel® Quark™ SE is the processor for the Intel® Curie™ Module, meant for IIoT.

Therefore, IIoT is relentless and continuously measures, monitors, makes decisions, and notifies the operators that something needs repair before it fails.

Scaling to this degree has not been possible until now, and companies like Intel and GE are teaming up to make one cohesive IIoT on levels starting at the IC chip in the sensors all the way up through the human that ultimately manages the outcomes that IIoT reports.

IIoT takes it to a whole other level, squeezing out the last drop of productivity and quality results.

Table 1: A partial list of benefits and concerns in implementing IIoT.

Table 1: A partial list of benefits and concerns in implementing IIoT.

Security

Security is such a concern that it needs further discussion. Security is already a problem for the millions already deployed, because many find it cheaper to send data wirelessly rather than install physical cables, as in “the olden days.” A grid might be taken down if the correct devices are hacked, and this is Cyber Warfare. But IoT and IIoT is cheaper to implement with CCTV images, for instance, since images can be sent wirelessly instead of installing cable. However, if robust passwords are not set upon install, or as in some cases, no password at all, it’s easy to use these cheap devices as for a Distributed Denial of Service (DDoS) attack, attack, and the owner of the device will not notice because it doesn’t take much code space. Ways to mitigate the IoT-device hacking include setting a unique password. Many people use passwords like “123,” which is easily hacked. Hackers run programs that check the passwords of Internet devices against dictionaries. Choose an IoT device with a processor with secure boot, which is chip-based and starts with the silicon. Secure firmware updates are then possible if the packet requesting access is authorized. Enable sending secure packets; use encryption. Control ports of entry to IoT devices. A connected car is considered IoT, and a USB port should only have rights to contact anything in the car’s media system, and nowhere else.

Eventually, the world will grow into the responsibility of IoT, which includes IIoT. It remains to be seen whether governments must legislate responsible IoT security within every IoT device (such as a complex and unique password by default, because they’re much harder to hack but better than nothing.) Nevertheless, the human race may have to learn the hard way by experiencing catastrophic failure. And it is doubtful that ubiquitous smart things have made humans smart enough to pre-empt an Internet disaster before it becomes a frequent announcement in the news as Internet-based companies experience data breaches and DDoS attacks that shut down sales and operations for hours or days. Sometime in the future as an incredulous older generation passes away, those who were raised on electronics from babyhood may finally set things straight.

__________________________________________________________________________________________________________________

LynnetteReese_115Lynnette Reese is Executive Editor, Embedded Intel Solutions, and has been working in various roles as an electrical engineer for over two decades. She is interested in open source software and hardware, the maker movement, and in increasing the number of women working in STEM so she has a greater chance of talking about something other than football at the water cooler.

Share and Enjoy:
  • Digg
  • Sphinn
  • del.icio.us
  • Facebook
  • Mixx
  • Google
  • TwitThis