Don’t hold your breath for the industrial IoT platform

10 Comments

The consumer Internet of Things may be hyper-competitive, but at least it’s easy to define. The average end-user wants relatively few things of their IoT right now: a smart home, a smart car, and maybe a bit of healthcare information. If you’re selling a hot new security or energy management or fitness app or device, there are only so many places to look for partners. But what about all the other “things?”

The industrial IoT will eventually eclipse consumer markets, in terms of both the number of connected devices and the volume and value of connections. But the market’s potential is so large because it’s not just one market. The only thing salinity sensors, tire pressure monitors and nuclear density gauges have in common is that they’re all very important devices that someone wants to connect (though not necessarily to one another). The industrial IoT is a nearly limitless collection of use cases, each with its own ecosystem of hardware, APIs, networks, security protocols and financial constraints. And unlike the consumer industry, which is forced by markets toward openness and standardization, it’s perfectly reasonable to expect that in many industries, vendors will create and maintain proprietary standards for the foreseeable future. All of that adds up to a mess for the enterprise, device manufacturer, or ISV that wants to get the jump on their competition and lead the charge.

On the consumer side, there’s hope. As Gigaom Research Analyst Adam Lesser pointed out, we’re making substantial progress toward a platform for the connected home. At the very least, the smart home is seeing some coalescence around a few key vendors, thanks to kingmakers like Google, Home Depot, and Lowe’s (with Apple in the wings). Plenty of standards and relationships are still up in the air (Will BLE and Wi-Fi kill Zigbee and Z-Wave?), but we’ve surfaced a handful of companies for which building applications and hardware is a reasonably safe bet. Due to a lack of dominant standards and hubs, the wellness and healthcare markets aren’t nearly as far along, but regulation and consumer demand should winnow down the field soon enough.

Industrial and enterprise development is far more chaotic. Without a broad financial incentive for a consolidated standards push, early adopters are largely on their own. At our October Structure Connect conference, I spoke with the CTO of a regional shipping company integrating real-time payload weight, stock levels, and relative positioning of other fleet vehicles. He admitted that he left many of the integration decisions up to his hardware providers — some of which had never developed software. He knew it wasn’t a long-term strategy, but “getting a jump on the competition was more important than building something that will last for ten years.”

That’s a risky strategy, but he was right about one thing: the benefits of sensor-driven data in the enterprise are too big to ignore. But while we’re seeing fantastic progress from groups like the IIC and OIC, and regulatory bodies are beginning to participate in the discussion, the near term will require a certain amount of in-house integration and creativity. That’s why we’ve decided to focus a large part of our research in the coming year on how businesses can best navigate the path toward connected intelligence while the dust settles on standards. In the first quarter of 2015, Gigaom Research will build on our previous research by examining the following:

  • Which devices to sensor-enable, and which sensors to choose
  • How to build and certify ecosystems and APIs around sensor-driven products and apps
  • Which different networks and protocols are available, and how to choose the right one(s) for the job
  • How to balance security concerns with time-to-market

Are there any issues we missed, or topics you’d like us to cover? Let us know.

10 Comments

Liz Presson

I definitely understand some of the comments on this piece, but I also think that some of the best points of the article are overlooked by rebuttals. (And let’s be honest, the headline makes it hard not to get in rebuttal mode.)

This statement is so incredibly spot on: “The industrial IoT will eventually eclipse consumer markets, in terms of both the number of connected devices and the volume and value of connections. But the market’s potential is so large because it’s not just one market.”

At Digi, every Industrial IoT solution we’re a part of takes different technology and a different networking approach. Our CMO, Jeff Liebl, wrote a great response to this article with that sentiment in mind here: http://bit.ly/1DhN4vx

Audi

I have to say I disagree with the premise of the title of this post stating that Industrial IoT is far away especially as a comparison with home or personal IoT.

In home automation I do agree that there is currently a lack of standardization, or more precisely several competing emerging standards. At least when it comes to the PAN or local networking of devices within one household.

But the equivalent here to the industrial world is “industrial automation”, which has been around for decades and is VERY mature. There are several well adopted protocols which nearly all industrial equipment manufacturers have built into their systems and is present in even relatively old equipment. Modbus, OPC, Profibus are all routinely used to network various systems in power plants, municipal systems, and various other large systems. Each industry perhaps having some unique variation, but within an industry for the large part all compatible.

I hear a call quite often for a standard protocol for IoT even in the industrial space, but often this is coming from traditional enterprise IT professionals who take a quick look into industrial systems and see only a mess, then call for everyone to adopt a new standard such as MQTT driven by IBM a more enterprise IT focused company. What is not seen is the the entire discipline of automation control systems which is already there. One ironic point, one often used protocol, OPC, actually has its roots in Microsoft technology.

When it comes to connecting systems to the Internet and forming an IoT system out of industrial systems, I do agree there is no standard, but that is actually a good thing perhaps. The internet grew out of the ability to connect computer systems to one another so they could essentially all exchange information freely. The foundations of routing protocols first instantiated in ARPAnet then built upon by companies like Cisco focused on exchanging information, not necessarily authorization and security control mechanisms.

For industrial systems you in most cases don’t want an open network. Imagine if any device or person could connect to a power plant, this would lead to severer infrastructure vulnerabilities on the scale that could cause severe economic harm or personal safety. For these types of systems, you really need to control what and who they can send data to, and who can control them. Thus rather than an open network where all nodes are connected to one another, what is required is a limited connection to one or a very few hub’s. From there they data can be controlled and monitored to be only given to the select few who are trusted to be able to monitor that equipment. So therefore having a single standard to connect to the hub’s is not needed and in fact is perhaps not desirable. Thats not so say that without the appropriate security controls you should not use the infrastructure of the internet. The ubiquity of the internet infrastructure is what i allowing the growth from automation systems into Industrial IoT systems.

I have worked in Industrial IoT systems for many years including at GE and now for many other companies. I can say from personal experience that the Industrial IoT space is already here, its just not as visible perhaps as connecting your home heating system and lightbulbs is in the current consumer buzz around IoT.

David Houghton

Cormac – a provocative headline that appears to be generating a little angst. Fact is, the ‘Industrial IoT’ is merely a creative label for the natural extension of automation that pervades this space and continues to evolve based on the advances in technology. As such, the notion that there is a dormant market is not accurate. Clearly challenges exist – this is not a plug and play world and won’t be anytime soon. But that is not new news. Reality is, what we’re experiencing is a growing market that is accelerating given the reduced costs and increased capabilities, which are literally reshaping business models. There’s insurmountable opportunity in making dumb things smart.

David Houghton
Bright Wolf

Peter Fretty

There has actually been some significant progress on the IoT front, it just doesn’t have the same media sizzle as the smart home where appliances are talking to one another. However, when paired with solid data analytic tools and educated people to put the analysis into play, the industrial IoT platform has far more meaningful potential. Actually I would venture to say it’s this growing space that is truly fueling the continued growth and acceptance of data analytics.

Peter Fretty, IDG blogger working on behalf of SAS

Pete

Also, many consider “smart home” more ‘consumer’ Internet than ‘industrial’, which is why we have groups like OIC doing work apart from IIC.

Rick Bullotta

I would recommend that you spend more time researching the state of the industry and the realities of the industrial IoT before writing “link bait” titles like this. Our ThingWorx platform, which is the most comprehensive platform for the Industrial IoT, has been available for a couple of years and is in use in many years at nearly 300 of the world’s leading businesses, connecting millions of devices and machines. With our recent addition of the capabilities of the Axeda platform to our offering, we provide an unequalled top-to-bottom stack for creation of innovative IoT applications. We also embrace ubiquitous connectivity to just about anything – from the device layer to advanced analytics. We have over 65 partners with pre-integrated solutions and capabilities that extend the ThingWorx platform with domain-specific functionality or deep integration with other applications.

Solving the connectivity and integration challenges of the IoT does not (and cannot) require any type of universal standards, since there are many billions of devices already installed and connected. There will continue to be broad proliferation of many protocols, “standards” (domain and technical), and applications that will increase the chaos, not reduce it. To that end, we designed our platform to embrace this reality and provide a unifying layer for data, services, events, and security that delivers real customer value and dramatically reduces the time-to-market (and cost-to-market) for connected products and applications.

I would be happy to arrange a briefing to educate you on the capabilities of the platform and the unique and innovate ways our customers have applied it.

Rick Bullotta
CTO/Co-Founder
ThingWorx

Cormac Foster

Rick,

Over the next couple of quarters, we’re actually going to be looking directly at the kinds of services companies like yours provide, to paint a picture of what you really can hand over. It’s safe to say that developers don’t want to manage the plumbing of their IoT initiatives, so the desire is there. I’ll let you know when we have the reports queued up, and I’ll be sure that all the relevant analysts are aware of what you’re doing.

Thanks.

mike

@CormacFoster I agree with my industry colleague @RickBullotta.

There seems to be much that you are unaware of regarding the current state and momentum of industrial internet solutions. I am sure that a little investigation would help you build an informed perspective.

I would be happy to host you at GE’s Software Center and brief you on what we have developed (and deployed…to paying customers…generating real value). We can bring your understanding of Industrial IoT up to, and way beyond, your perspective on consumer IoT.

You might even understand why I respectfully disagree with your article’s conclusions regarding consumer IoT.

Mike Dolbec
Managing Director, Venture Capital
GE Software

David Friedman

Great article – but while it is very likely, and even desirable in many cases that industrial systems be separate and NOT communicate with one another, IoT platforms from leading vendors should span into the industrial space just as they do in the consumer space. I would imagine that many of the baseline requirements from the shipping company are shared by other participants in the industrial IoT: provide secure 2-way communication of data from sensors/”things,” enable messaging, rules, store the data, provide big data reporting and analytics, etc. The industrial IoT will of course have additional customization around the edges, will have more stringent security requirements… and many more different requirements and challenges. However, industrial IoT is very similar in that companies who are not experts in building these systems should definitely NOT be building them. This is not only as it relates to the inefficiency of each company trying to “recreate the wheel,” but much more importantly that the systems are not going to be secure or scalable if the architects and developers who build the systems are not deeply experienced. Moreover, we know that if a server is connected, it is prone to hacking and viruses, and participants in the industrial IoT need a team that is ready 24/7 to identify these attacks and thwart them. The market is definitely more fragmented than consumer, and will move slower, but it is very true that the value of the data is massive…

Pete

Great points. I’ve been in a few of these 24/7 monitoring centers and am also familiar with Axeda as well as Rick’s work, below, and the IIoT is definitely a bit further along for some than for others. I think the topics focused on above are on point, with maybe the addition of some explorations into the value chains between customers and service providers. Looking forward to what Gigaom uncovers!

Comments are closed.