4 Comments

Summary:

Get ready from some serious hype. Five of the largest companies touting the industrial internet have teamed up to form a standards organization to show us what connectivity and data can do.

What can a connected GE jet engine tell you?

The giants of the industrial internet and the federal government have banded together to form a standards organization to help make connecting sensors, building networks and sharing data just a bit easier. And when I say giants, I mean giants: Founding members GE, Intel, IBM, AT&T and Cisco have created the Industrial Internet Consortium under the Object Management Group.

That’s right, the IIC is under OMG. The effort looks like a commercial version of what Qualcomm and consumer brands are attempting with the All Seen Alliance under the Linux Foundation. However, I’m sure All Seen will not be thrilled about the prospect of playing a limited role as the consumer standard for the internet of things, giving up the industrial side to this new group.

This may look like a big deal, but most of these companies are already working together. AT&T, GE, Cisco and Intel announced a partnership last fall, and IBM has been pushing its MQTT standard since April. The most interesting element of this creation isn’t that it exists, but that it may provide us a view to see Cisco’s smart cities efforts come head to head with IBM’s Smarter Planet initiatives. Or whether GE’s investment in Pivotal wins out on the data warehouse and analytics side or Intel’s new championing of the Cloudera Hadoop distribution finds its way into the standards.

Like most standards groups, the goal here is to build interoperability into the gear and software that enterprises will use to collect, store and analyze data coming from their connected devices (and power grids and roads and cars, Oh my!). Membership is open to all, and the initial focus will be around developing scenarios, reference architectures, and business processes for the industrial internet.

Yet, given the sheer amount of grandstanding in the press release about the epic change that the industrial internet will cause, I’m going to wait for the IIC to actually do something. In general, the amount of good a standards organization will do can be measured as the inverse of the puffery in the release. I’ll also note that the release introduces a new word into the lexicon: cyberphysical.

This appears to be the result of the government’s involvement (the release has a quote from Penny Pritzker, the Secretary of Commerce) and the release touts that the government is investing “over $100 million/year in R&D related to “cyberphysical” systems, and has been partnering with the private sector on a series of testbeds in areas such as healthcare, transportation, smart cities, and increasing the security of the electric grid.”

Here are some things taken from the IIC web site and release just to give you a flavor of this new utopian future that awaits after we deploy the magic of the industrial internet to create a new “cyberphysical” reality:

  • “Predictive policing through video analytics that can pinpoint likely crimes before they happen, and to deploy emergency response resources on demand” — from the public policy scenario web page.
  • Create transportation systems that can sense and respond to changes in real time — from the transportation scenario section of the web site.
  • “We are at the precipice of a major technological shift at the intersection of the cyber and physical worlds, one with broad implications that will lead to substantial benefits, not just for any one organization, but for humanity,” said Janos Sztipanovits, E. Bronson Ingram Distinguished Professor of Engineering and Director of the Institute for Software Integrated Systems (ISIS), Vanderbilt University.
  • “Ninety-nine percent of everything is still unconnected…,” said Guido Jouret, vice president of Internet of Things Business Group for Cisco.

To be clear, I too believe that adding real-time information and connectivity in more places will change businesses. I’m not sure it will generate the “new jobs” that Pritzker references in her quote, or if those jobs will be ones that people will consider good. I do however think that most enterprises will end up deploying a custom and deeply integrated set of hardware, software and services to truly change their businesses for a connected world, so this idea of a standards organization that will make that significantly easier rings a bit false.

But maybe the masters of marketing at each of these companies felt like the “cyberphysical” industrial internet of everything needed a bit of a public relations nudge. It’s going to save the world y’all.

  1. Interested mixture of companies in the arena. Indeed with existing and competing agendas, it is very hard to see if this is going to help or hinder in this environment. On the other side, this is a tremendously complicated space and its good to see a movement to standardization.

    BTW: Cyber-physical has been around for a while as a tem. http://cps-vo.org/

    Share
  2. The Internet of Things (IoT) is diverse; covering process plant control devices, discrete manufacturing automation devices, home and office automation devices including all appliances, in vehicle components like engine controls, transportation infrastructure including all traffic signals, railroad signals, and rolling stock as well as utilities infrastructure including the power, water, and gas grids metering etc. The devices at the lowest level of these systems are mostly sensors and actuators. These devices are different from the devices which serve the “Internet of People”: user interfaces such as computers, tablets, and smart phones etc. which is much more homogenous. Fortunately digitization of sensors and actuators in the various fields has already begun several years ago and many of the required digital communication protocols for these digital devices already exist. It started with many proprietary protocols, most of which have now been eliminated. It is crystallizing to fewer, yet several standard protocols, but this is because these automation areas have very diverse needs so one protocol cannot serve them all. Therefore we see different protocols in each area: FOUNDATION fieldbus for process control, PROFIBUS for motor controls, SERCOS for motion control, EIB for building automation, CAN inside cars and other vehicles, a.s.o. These have not only application protocol differences, but also physical layer signaling differences because they must cover diverse and conflicting requirements for distance, speed/bandwidth, response time / determinism, bus/loop power, intrinsic safety, and topology etc. You can’t replace all with a single protocol or media standard. So the IoT is about interfacing the lower level network protocols from sensors and actuators at the very “first meter” through aggregating devices like controllers, up through the enterprise, onto the Internet through a secure connection where they can be accessed from anywhere. This is just like the Internet of People where it doesn’t matter if your keyboard, mouse, webcam, or other ‘sensors’ uses USB, Bluetooth, or RFID etc. – people on the other side of the world can still see you and get your messages.

    I personally think the main challenge at the moment is that most of the industry is quite far behind. Most plants today use sensors and actuators using analog 4-20 mA and discrete on/off signals. The very first step is to get plants digitized; new plants have to be built using real-time digital communication standard such as FOUNDATION fieldbus and PROFIBUS (IEC 61158 family), and WirelessHART (IEC 62591) instead of 4-20 mA and on/off signals etc.:
    http://www.fieldbus.org/images/stories/technology/aboutthetechology/overview/fieldbus_brochure.pdf

    Developing the first version of these standard fieldbus protocols with interOPERability was easy. However, the enhancements with interCHANGEability also making the systems easy to run & maintain was more difficult and took 20 years – we are only just reaching this goal. The demands of the process industry is that a digital device can be replaced using only a screwdriver, without clicking software, just like for a 4-20 mA device. This is now possible with fieldbus, thanks to the Electronic Device Description Language (EDDL, IEC 61804-3, http://www.eddl.org/) and device profiles, but still cannot be done for a printer in the home/office world so the automation industry is far ahead of the office world with respect to interchangeability. Simple USB devices are not too bad though. So to get the Internet of Things off the ground in a timely manner, I believe we must find ways to at the lowest levels use and integrate the existing protocol standards up to the higher levels and through a secure connection to the Internet.

    Interchangeability is required for easy maintenance of the IoT, but interchangeability is more difficult than interoperability. It took a long time to achieve for fieldbus.

    At the intersection of the cyber and physical worlds we find sensors and actuators. The main barrier today is not that these are in different technology silos, the problem is that they are not digital, a vast majority are still analog and discrete. The first order of business in my personal opinion is therefore to digitize applications, instead using digital sensors, based on standard protocols such as IEC 61158 and IEC 52591. Using such standard will diminish the need for custom software development.

    Share
  3. Angelo Corsaro Tuesday, April 1, 2014

    Nice article Stacey. That said, I’d like to point your attention to another standard technology that is used in Industrial Internet (I2) Applications: the OMG Data Distribution Service.

    DDS is today at the foundation of several IoT and I2 systems that have been deployed and operating for some time. One of the best examples is the Nice (France) smart city infrastructure (see http://www.youtube.com/watch?v=neVyOTXB4eI).

    The main strenght of DDS is the technology itself (you can learn more at: http://www.slideshare.net/angelo.corsaro and http://www.prismtech.com), especially when compared with the available alternatives.

    Its weakness, if we can say so, is that the standard is relatively new (10 years younger than MQTT) and mostly implemented by innovative small/medium companies.

    I think it would be great if you could help create awareness of this great technology which, regardless of the limited buzz, is used in real systems and has a huge potential to really help building Industrial Internet applications.

    – Angelo

    Share
  4. Why is it that every time these companies get together to develop a new “standard”, we end up with a hodge-podge of bifurcated implementations?

    None complete, no one “standard”… and a mess on the other end trying to glue it all together.

    Share

Comments have been disabled for this post