The Industrial Internet of Things


Architectural principles of I2oT:

  • ISA100.15 illustrates methods for segregating automation networks in zones and connecting the zones with conduits. The concepts for zones and conduits were developed by ISA99 for ISA/IEC 62443.
  • ISA100.15 illustrates methods for creating and enforcing policies that determine which entities can communicate over industrial networks, and the allocation of bandwidth or relative priority of those communications.
  • As a general principle, many automation nodes and networks have limited bandwidth and capability to serve data. The primary source of data from these nodes must be in separate buffers and caches with sufficient bandwidth for expected users. Direct access to nodes requires systematic limitation.
  • Local autonomous control and local history collection and compression are two techniques that are used in industrial automation systems to deal with low bandwidth and/or high or variable latency in networks. Many proprietary schemes are used to fill these needs, and there is an opportunity to standardize these features.

Common network management—ISA100.20 is starting work on a recognized need for common network management. This is a gap that needs to be filled. Common network management is intended to provide a way of managing multiple diverse networks with common tools. Some of the tools already exist, but many networks are not designed to work with external tools. Some networks are unmanaged. This is an opportunity.

Common security management—This aspect will also need work. It is not easy to separate this need from common network management, and it may be handled by the same effort. This will need to include provisioning procedures and tools. Most industrial protocols do not include sufficient security. They need to be upgraded to include this as a standard feature.

Specifications and profiles to support compliance certification—We need plug and play, not plug and pray. This means that we will need enough compliance profiles and specifications that a certification body can do compliance testing. Standards developed by a standards development organization may support this cause, but will not be sufficient to meet industry needs.

What does the future hold?

There are multiple paths and even multiple end points for this technology migration. It is not certain what path(s) or endpoint(s) will be the final result. The only certainty is that this technology will change, and so far the changes have led to more diversity.

Current state—Many organizations are involved in pieces of the communication problem, but no single organization is involved in all aspects of this problem and no organization is well positioned to take the lead in coordinating this technology drift.

Some standards organizations have taken on the task of standardizing horizontal layers of the communication stack. IEE writes standards for the PHY/MAC (or link) layers. IETF writes standards for network and transport layers. Coordination at the boundary between the link and network layers (which is also an organizational boundary) is informal at best and sometimes appears to be a turf war.

Many foundations have taken a vertical approach to communication stack specifications and have written full top-to-bottom stacks. Many of these support only one PHY and one application and have a single stack in between. Others have multiple PHY layers available but with restricted stacks in between. These foundations are generally market competitors.

All of the organizations fill a need whether they are working on a horizontal or vertical slice of the communication problem, or just a little piece of the problem. All of these organizations could slowly converge on best-in-class technology, or not. If market forces do eventually drive this convergence, it is safe to assume that the convergence will happen very slowly and chaotically under market conditions.

Lead organization—Building a positive outcome will need a lead organization or a consortium of organizations to make this effort successful. It will involve liaison work with too many organizations to name, and will need executive sponsorship from multiple entities including vendors, standards organizations, and foundations.

Herman Storey is chief technology officer of Herman Storey Consulting, LLC.

See additional voices on the topic on the next two pages:

Anonymous , 08/06/15 02:37 PM:

In my personal opinion, based on experience, the less disruptive we can make IIoT the faster it will be adopted. It is when disruption occurs that adoption is delayed (read “Crossing the Chasm”). For instance, if the lines between Instrumentation & Control (I&C) side and the IT side have to be blurred, there becomes much confusion and apprehension. If the lines of responsibility can be kept crisp, and where each has minimal dependency or interference on the other, the adoption will be faster. The same goes for the lines of responsibility between the instrument group and the control system group. If there is a clear line of responsibility between all then each group knows what to do to make the information flow unobstructed. For example, an instrument technician must be able to replace a transmitter in the middle of the night without consulting with the IT department and without getting help from a system engineer. Technology can evolve but it is more difficult to change established ways of doing things around the plant. Both business systems and control systems can use Ethernet and TCP/IP technology, but the way these two systems will be managed is different due to mutually exclusive requirements. Therefore, let each department manage their own infrastructure the way required by each application – they meet at a single well defined interface point. Avoid mixing IT network and I&C networks in the same box to enable each department to manage the network to their requirements. Just because the technology is the same doesn’t mean it has to be managed the same way. One might think that common management brings some benefits, but if there are more drawbacks then keeping them separate makes sense.

I personally agree with Herman that a single physical medium will not serve all requirements. Just like our home and office need both Ethernet and USB, and both Wi-Fi and Bluetooth, the plant environment needs both Ethernet and fieldbus, and Wi-Fi and WirelessHART.

IPv6 is not necessary for IIoT. What is required, by definition, is that each device (‘thing’) has a unique identifier. An IPv6 address is one way of uniquely identifying a device. FOUNDATION fieldbus and WirelessHART and other industrial protocols have unique IDs that work just as well. Plants built on these technologies are in a good position to move to IIoT when the time is right.

I agree that many industrial protocols already exist in plants connecting millions of devices. These devices lay the foundation for the IIoT so any developments in IIoT need to be backwards compatible with these devices and networks to ensure speedy acceptance of IIoT paradigm where these devices are accessed remotely from other offices.

What is required to make data flow freely from sensors to the enterprise is communication without barriers. This requires that the application protocol which is used over IP networks is the same as the fieldbus application protocol used by devices that sit on automation networks like FOUNDATION fieldbus, WirelessHART, Modbus/RTU, PROFIBUS-DP, and DeviceNet etc. That is, the corresponding application layer protocol used across Ethernet and other IP network are FF-HSE, HART-IP, Modbus/TCP, PROFINET, and EtherNet/IP respectively. By using the underlying fieldbus protocol with the corresponding Ethernet protocol the data can flow freely and unobstructed without gateways, drivers, and data mapping etc. A simple linking device is all that is required to connect Ethernet to a fieldbus. This in turn enables intelligent device management (IDM) remotely:

Upgrading fieldbus protocols to use IP at device level would mean all the fieldbus protocols would have to be changed. Manufacturers would have to maintain two concurrent versions of the fieldbus device (in addition to the 4-20 mA/HART and WirelessHART versions) since existing plants already use these. At the instrument level IP adds little or no value (since devices already are uniquely identified and can be accessed remotely through an IP linking device) so it may not make sense to change all these protocols creating new versions of all of them.

I agree we should use the existing application layer protocols rather than creating new protocols. Indeed it is possible to create IP devices that support multiple application protocols. For instance one application communicates with the device using HART-IP, another application uses FF-HSE, a third application using PROFINET, and a fourth application using Modbus/TCP – all from the same device, all applications getting the same data but through a different IP port. This requires the multi-protocol solution is implemented as intended, like on Ethernet and Wi-Fi, where the application protocols access the device directly through their assigned IP ports, not by “tunneling” across another application protocol of some type.
Anonymous , 08/06/15 02:38 PM:

I personally agree with Rick that backwards compatibility and integration with existing network technologies and devices is important. These devices are the foundation of IIoT. There are linking devices available for connecting all fieldbuses to IP-based networks: WirelessHART to HART-IP, FOUNDATION fieldbus H1 to FF-HSE, Modbus/RTU to Modbus/TCP, PROFIBUS-DP to PROFINET, and DeviceNet to EtherNet/IP. The same application protocol is used on the IP side as on the fieldbus side meaning data flows freely without barriers such as data mapping.

The way whereby devices are identified to humans in the industry is by “tag” name. So device tag is the logical way whereby devices shall located in the IIoT. Existing industrial networks like FOUNDATION fieldbus and WirelessHART already support locating devices using device tag. That is, a device is identified as for example 51-PT-101; humans do not need to know the address of the device. These protocols are the foundation of the IIoT. Plants built on these technologies are in a good position to move to IIoT when the time is right.

Modern protocols like FOUNDATION fieldbus already support communication of real-time and non-real-time data in separate timeslots such that non-real-time data does not delay real-time data.
Anonymous , 08/06/15 02:38 PM:

I personally agree with Drolet that existing industrial networks cannot be replaced by IP-based communication. Instead, continue use of fieldbuses, connected to higher level Ethernet and IP through linking devices. A key element is that the Ethernet application protocols be the same as the application protocols for the underlying fieldbuses: Modbus/RTU<>Modbus/TCP, DeviceNet<>EtherNet/IP, PROFIBUS<>PROFINET, FOUNDATION fieldbus H1<>FF-HSE, and WirelessHART<>HART-IP. This ensures data flows unobstructed without undue integration effort.
Consulting-Specifying Engineer's Product of the Year (POY) contest is the premier award for new products in the HVAC, fire, electrical, and...
Consulting-Specifying Engineer magazine is dedicated to encouraging and recognizing the most talented young individuals...
The MEP Giants program lists the top mechanical, electrical, plumbing, and fire protection engineering firms in the United States.
Salary survey: How much are you worth?; Dedicated outdoor air systems; Energy models and lighting
Fire, life safety in schools; Fire protection codes; Detection, suppression, and notification; 2015 Commissioning Giants; Emergency and standby power in hospitals
HVAC and building envelope: Efficient, effective systems; Designing fire sprinkler systems; Wireless controls in buildings; 2015 Product of the Year winners
Designing positive-energy buildings; Ensuring power quality; Complying with NFPA 110; Minimizing arc flash hazards
Implementing microgrids: Controlling campus power generation; Understanding cogeneration systems; Evaluating UPS system efficiency; Driving data center PUE, efficiency
Optimizing genset sizing; How the Internet of Things affects the data center; Increasing transformer efficiency; Standby vs. emergency power in mission critical facilities
As brand protection manager for Eaton’s Electrical Sector, Tom Grace oversees counterfeit awareness...
Amara Rozgus is chief editor and content manager of Consulting-Specifier Engineer magazine.
IEEE power industry experts bring their combined experience in the electrical power industry...
Michael Heinsdorf, P.E., LEED AP, CDT is an Engineering Specification Writer at ARCOM MasterSpec.