18 November 2020
Unify or Uniformize your I(o)T?
The need for unity runs through every organization, whether it's society, family, sport (collective, though not only) or business. Unity lets a group behave coherently and efficiently, moving toward a common goal rather than acting as a series of individuals. A physicist would say it reduces, at least locally, the entropy of a system.
The temptation of uniformity
To reach unity, it's tempting to approach it through uniformity. If everything looks the same, coherent collective behaviour is obviously easier to achieve: the same uniform, the same jersey, the same housing, the same reading.
And it works, up to a point. Totalitarian societies understood this well, imposing strong social rules to control their citizens and avoid being questioned. On a lighter note, common conventions are genuinely useful: the same power outlet, at the same voltage and frequency, or shared codes (blue means cold, red means hot; green means go, red means stop).
The real question is how far it pays to standardize. At what level should rules apply, and how much flexibility should individuals keep?
There's no obvious answer, and it depends on context. Standardization pushed too far into the detail (over-specification, over-control, over-definition) produces a rigid organization with little trust, little room for innovation, and poor adaptation to change. Not pushed far enough, and the system risks misalignment and misunderstanding between its parts.
To unify is to communicate
Unifying a whole requires good communication between its parts. In a uniform world, that communication is implicit: rules carry most of it, and that's enough in practice. In a more flexible, creative environment, communication has to happen at a higher level. In team management, this is the difference between managing by process and managing by objectives.
The second approach is more rewarding for participants and usually delivers better results, but it demands regular exchanges between team members to keep everyone aligned on shared goals: reading the manual doesn't solve every situation.
What about I(o)T?
Unify or uniformize, that's the question behind many IT and IoT projects. Is it better to pick a single solution with broad functional coverage from a dominant vendor, accepting whatever limits come with it, now and later? Or to pick the best tool for each function and take on the integration work needed to deliver the expected service? Mainframe or distributed workstations? A generalist ERP suite or specialized software? Company phones and laptops, or BYOD?
There's no single answer: the context, and the culture, of each organization dictates the response. That said, whatever level of standardization is chosen, there's always a need to integrate different products, services and objects. Two areas make this particularly visible.
Applications, first: it's telling that they aren't standardized. At best, some exchange formats get specified, or imposed, by a dominant vendor, not without friction. The reason lies in designers' imagination, ever-shorter product cycles, fast-moving technology and the drive to differentiate: no vendor wants their product called a copy.
IoT, second: the sheer diversity of devices, the number of network technologies, the proliferation of platforms and the applications they must talk to make uniformizing large projects nearly impossible, or at least unreasonable. Unlike the world of applications, IoT doesn't suffer from a lack of standards but from an unreasonable abundance of them: the standards bodies trying to fix this deserve credit, and face plenty of difficulty.
Unity without uniformity needs a common language
APIs are the entry point to the different components of an IT system, exchanging commands and data. Unifying an I(o)T system around business processes requires the APIs of its component applications to be used consistently: a genuine conducting job, where the instruments differ from one another and don't always share the same musical rules of key, rhythm or tone.
The task quickly becomes unmanageable as the number of instruments grows and the score gets more complex. To avoid cacophony, the players need to share at least one common language.
Users shouldn't be left out either. They have two defining traits: they aren't uniform, and don't share the same goals, moods or level of knowledge; but they do share a common language, our natural language.
So isn't it tempting to use natural language as the communication layer that unifies connected objects, applications and users, in service of business processes that stay flexible, unified and not too uniform, and remain open to innovation?
Want to know more?
Request a demo and see how Agora can transform how you manage your assets.
Request a demo