Industrial IoT platforms: Is acquisition the path to completeness?

The IIoT platform market has barely got to the wobbly toddler stage.  And yet, this week has already seen some high profile M&A (actually, just A...) activity.  Greg Gorbach and I have cogitated quite a bit lately on what defines an enterprise Industrial Internet of Things platform.  We're about to put pen to paper on a strategy report in this area, so here's a glimpse into our thinking so far.

When it comes down to it, in our view there are 8 broad capabilities that a well-rounded platform for the Industrial Internet of Things needs to have.  In no particular order:

1. Device interface and management.  IIoT data has to come from somewhere, and a lot of it will come from various sensors attached to devices, and through gateways and device connectivity platforms.  An IIoT platform will need a way to establish and manage connections to those devices, and collect and route the data from those devices.


2. Data management/processing engine.  That IIoT data obviously has to go somewhere, and it will go into some type of database - or two.  Expect platforms to be capable of ingesting lots of fast moving data, storing it, and providing a foundation for applications and analytics.


3. Application enablement.  All that data has little value without either apps of analytics.  Expect application building environments to be agile and follow the devops approach, and also to create microservices in addition to complete applications.  Most likely, IIoT platforms will be cloud-centric, so expect a Platform-as-a-Service approach to application building.


4. Analytics.  If it's not IIoT apps, it's IIoT analytics - and a complete enterprise IIoT platform should have both.  The range of analytics that may be required is quite extensive - from high performance analytics of structured data, through real-time analytics of streaming data, dealing with time-series data, embracing machine learning,  and on to include unstructured data such as video or audio.


5. Operational intelligence/business processes.  Simple operational intelligence - such as routine monitoring and alerts when some condition arises - will be required in addition to the more extensive analytics described above.  Rules engines to automate decisions and workflow can help people to focus on what's important, and also ensure that time critical issues are addressed promptly.


6. Enterprise integration.  Quite apart from device connectivity, other data integration may be required too - for example, to integrate asset maintenance records held in a relational database with sensor data being generated by those assets.  Beyond source data, integration will also by necessary at the other end of the solution.  Analytic insights, apps, and microservices developed within the platform may all be embedded within existing enterprise applications and workflows.


7. Simulation and modeling.  Certain applications benefit from the capability to simulate IoT devices.  For example, by creating a 'digital twin' of a machine, then comparing its behavior to real-time data from the physical machine, it may be possible to predict failures in advance.


8. Security.  Robust and pervasive security is required from top to bottom - or inside to out, depending on how you want to think about it.  For example, at the device level it's critical to ensure that only authorized devices are able to access the platform.  Likewise, at the user level, analytics and apps should only be surfaced to authorized users and devices.  And in between, data needs to be secured both at-rest and in-transit.

Those 8 broad categories of capabilities provide context for some of the announcements that we've seen this week.  For example:

• Cisco announced the intention to acquire Parstream.  (Cisco's blogged announcement here, the more acerbic take of El Reg here).  Cisco makes no bones that the acquisition is about analytics at the edge.  I'd probably go a bit further than that and say that it's about analytics period.  In other words, Parstream will strengthen Cisco in one of the essential areas required of an IIoT platform.


• Intel buys Saffron Technology.  Saffron is perhaps not an obvious Industrial IoT player, but it has plenty of potential there in my eyes.  Marketed more as a cognitive computing solution in the vein of IBM Watson, I see this as another analytics acquisition as part of an IIoT platform play.


• Jasper will partner with IBM to round out the device connectivity options for Bluemix.  IBM Bluemix is a horizontal Platform-as-a-Service solution.  There is a Bluemix IoT variant which is part device specific templates, and part curation of existing services.  The partnership with Jasper promises much stronger device connectivity and management.


• Verizon is launching ThingSpace,  a new IoT platform allowing developers to create applications, customers to manage devices, partners to market their services, and Verizon to launch integrated vertical solutions simply, in an open environment.


• IBM acquires the The Weather Company's product and technology.  To American's at least, The Weather Company is best know for it's all pervasive Weather Channel.  IBM had announced a partnership with The Weather Company earlier in the year.  But, an acquisition..?  Have to admit, I did not see this one coming...Here's where it gets interesting from an IIoT perspective.  Turns out The Weather Company has built out a very impressive infrastructure for collecting and working with IoT data.  For example, over 40 million smartphone users have The Weather Company's app installed - and the majority allow the app to use their phone to sample barometric pressure and report it back.  (Editor’s note:  I wonder how many of them actually know they gave permission for that…)  In addition, The Weather Company ingests data from 140,000 personal weather stations worldwide, as well as planes and cars.  Combine that with over 151 forecast models, and the company has over 2.2 billion specific forecast locations, updated every 15 minutes, with a total of up to 26 billion forecasts a day.  And IBM just bought the technology infrastructure that unpins it.

So that's it.  One short week for IIoT platforms, yet 3 notable acquisitions, one interesting partnership and one major product launch.