Value and Scale for IoE. How do we get our arms around that challenge...

In the last 2 weeks there have been a couple of superb opportunities to share and discuss views on how to shape up scalability of the Internet of Everything and rethink value creation.

Last week, I had the opportunity to hold a talk at IoTCON in Berlin together with my friend #Thomas Eichst?dt-Engelen (one of the lead figures of #Eclipse Smart Home and #openHAB, and I should acknowledge that I stole two charts from him for this article...), while this week we had a great session with #W3C and industry partners at #IndustryofThings World, and where invited to be on of the “hottest 15 IoT startups” at the same conference (actually we were only 14). 

Next, I am just travelling to meet the #tmforum guys on IoE monetization in London, and we probably will have many of the same discussions in this setting again.  It is all around the question, how to deliver on the promise that IoE will be at the core of 25-30% of global GDP in only 10 years time…

 That is a cool thing if it would ever happen.  The key promise is that most of the value is actually not in the things, but how things do connect to the internet…let’s look at this. 

We have this great example here of a plug.  That plug may be remotely switchable, and it may even report on the energy usage.  The truth is that there is always some more challenge deep down, but still yes, that is great.  And there is an app to that. 

Cool, because the use case is super clear:  I am in my armchair, 3 meters out of reach of the lamps switch, and let’s wait, there was my smart phone somewhere, I unlock it in the dark, find the wemo app, and … boom, the light goes on…would have been even cooler, if the lights were already on, when I searched for the phone, but, hey…

When I listened to the presentations of the other hot IoT startups at Industry of Things World today in Berlin, I was so excited.  The quality and professionality was mindboggling.  There even one where Andy von Berchtolsheim invested and one of my dearest former McKinsey collegues, Lothar Stein gave reference.  There were incredible success stories that a couple of years ago would never have been delivered by startups. Cost savings for remote maintenance at Deutsche Bahn (#KONUX), at Oil and Gas companies, monitoring solutions for the manufacturing industry (#dat-uh, #indegy, #teep trak), great tools for energy management in breweries and all kinds of process industries (#factor-E),…But in essence, 12 out of 14 solve great use cases (the other more platform startup was actually #m2mgo, but still creating their kind of silo, but a cool piece anyways), and at the same time will create great siloes for the future.  So, great headache ahead.  While it is cool to switch on the light in the dark with your smartphone that was somewhere, let`s wait a second…, 50 of them definitely will create a problem. 

You may think, I do not know what I am talking about…but face reality...every household today has an average of 50 connectable – or even worse – already connected devices.  And when it comes to industry, the picture get’s even more dramatic.  The good message:  The proposed solutions still deliver a great benefit over manual performance management, maintenance and monitoring solutions in use today.  There are statements, that even siloed solutions can reduce idle time by 20-30%, maintenance cost by 25%, and reduce energy consumption conssiderably.  So, they hopefully all pay their cost.

But, they are not yet part of a sustainable IoE strategy. That is acknowledged by the industry.  The master key word is therefore interoperability.  That is cool.  First we had tens of competing technologies and proprietary solutions, now we have 35 initiatives to create interoperability, and counting.  That definitely will help us solving the problem. Sometimes, some of them declare victory. In the industry space today it is OPC-UA.  In the consumer space it is OCF.  Let’s see. 

Here is my dirty, little conviction…it is not enough.  You need to solve the magic triangle of IoE:  Security, connectivity, interoperability.  Yes.  On security, it is all about creating local trust in the end points, apply state of the art internet security on transport layer, storage, identification and authentication, as well as ever smarter algorithms to detect anomalies (while keeping out NSA).  On connectivity, there is a new standard every day.  That’s ok, as long as there is a good reason for it.

On interoperability, we will bang our heads, on how to connect all the dots between all the interoperability consortia, and how to keep with all the emerging classes and data models that are contributed by all the members into those data models. 

Here comes the good news:  Whoever will provide a solution to that in the end, it is already clear, how this solution will look like.  We already learned that in school.  There is always a biggest common denominator, as there is a smallest common multiple.  The latter solution will either be a fantastic super data model and specification database in which nobody will find anything that he doesn’t already know.  So, that is the nightmare solution, but nevertheless, very close to what we have today. 

The alternative is a smarter one…all things share some basic capabilities.  They can either be set or written on, or they report values, or you can read from them.  That are the context free basic capabilities. 

Then there is more complex things that help us understand the machine world as humans:  We need to understand the units of measure to understand the context of a value, and we need to recognize the format of what we read and write to storages.  We need further be able to classify the kind of connectivity, security and details on technology, as well as the manufacturer of a certain thing. 

Dependent on the concrete situation, we want to have additional information:  Relationships to other things, groupings, rules, location, ownership,…all that is context information that should all be organized in a modular way. 

Units of measure and other format information will often be linkable to existing standards,  like SI units or an eTOM process model, or a standard convention on universal postal addresses, or user information.  There will be the need for distinct modules on connection, security, technology as well as manufacturer info.  You can build all kinds of complex thing from those basic concepts.  A ceiling lamp is the probably a thing that is composed from several component lamps, that have the capabilities that they can be switched, dimmed and colored.  Thos capabilities are built up by combining basic capabilities with relevant context, e.g. units of measure (Boolean, % illumination, color,…). 

Most of the additional useful context information is actually not in the technology domain, but will be added by the user, or pre-configured by the application provider.  So, there is a quite natural way to build up information models. 

The nice thing about this approach is that it allows to easily map all existing models.  They all rely on basic capabilities – they need to, because otherwise they cannot talk to the machines they monitor and control, and they added context in some shape of form that is implicitly or explicitly build up from the same modules we were just talking about.

So, a universal representation model for IoE is either overly complex, or looks exactly as outlined above.       

If things are so clear, you simply need to make them happen.  So we did…

Here is our wonderful middleware with its main components, as we have them today…we represent things that we add to our platform with connectors that translate the underlying technology to our modular structure.  A user has a unit (think of it as a container) in which he has his things instantiated.  If he is the owner of that unit, he can add and remove things and services as he may wish, and he can delegate access rights and assign roles to other users.  He can create relationships between things (e.g. groupings) and can define rules.  He can do that  high level, or go down to any level of detail, e.g. individual properties and action, and assign rules and rights there. 

Technologies either get added cloud to cloud or they can be added by using our “thing API” and SDKs to directly link up technologies. 

In the end, all this nice stuff is to realize cool use cases.  But you do not need to do it this way.  We rather encourage to turn it around. 

The next 2 days, I will spend on IoE monetization with the tmforum folks in London.  We will talk a lot about platform capitalism and things like that.  That is at the core of all that.  But we will also definitely talk about the following little thought experiment: 

Take the magic triangle, and assume:  Security, connectivity and interoperability are solved.  So it is all about the creativity of the developer.  Think about the big opportunities in transport, industrial manufacturing, logistics, buildings, city management, healthcare, or even product development, organizational design and process automation. 

Where are the 10x opportunities, if you remove all barriers.  That is your list of priorities.  If there is 10x in value, productivity, quality, lifestyle, … just go after the use case.  There is enough room to fit in the right connectivity and security and safeguard interoperability.  Consider it solved. That is, what we need to do first thing!

And – by the way – the connctd team will be very happy to bring that to life.

The guys even now started a new cool feature…it is very much in ideation stage, but we couldn’t hold backfrom exposing it to you to gather feedback early on…we made capabilities searchable.  It is all there on api.connctd.com. Start your world changing project today. And give the team some encouraging feedback.