Industrial IoT: Convergence of two divergent worlds

The world is overwhelmed by technology at the moment. The disquietude of machines taking over humans, the aspirations to create revolutionary technology and the mandates to catapult businesses to unprecedented growth is all pervasive. Achievable or not, only time will tell. Amidst this highly entropic business environment, there are few technologies, or buzz words rather, that are gaining momentum. I say buzz words because companies around the globe are rechristening their offerings to fit these buzz words. ERP implementations are digital, analysis of data from factories is IoT and simple 'what-if' statements are AI. I'm sure many of you would have witnessed this already. But this does not mean we are not progressing, we are. Quite rapidly in fact. This means that multiple industries are coming together around these key buzz words. They are competing and they are collaborating. Coopetition seems to be the flavor of the decade. This is especially true for IoT, the acronym for Internet of Things.

IoT is not just about getting data from the physical world (motion, temperature, location etc.) and acting on it, rather it is about putting that on the internet. Looks obvious but not quite so in practice yet and highly important to understand if one cares about how to service one's clients. It's important to note that we are talking about 'Internet' of 'Things'. The two are fundamentally different ecosystems. Internet is the new born, agile, prefers to fail, adapt and believes in agility. The 'Things' world has existed for decades, cares about nothing more than 'quality' and does not launch anything that's not tested atleast thousand times.

The 'Things' world was created during the industrial revolution and has consistently been improving. It's a world where things not working fine is unpardonable. The workers, engineers, managers and leaders of this industry look at work in a very different way. If things are not manufactured right, billions are spent in calling the defected products back. This is a place where often six sigma fails to impress people. This is a highly process oriented, governed, hierarchical and regulated ecosystem. Partnerships exist for decades. Adam Smith's 'invisible hand' has worked its magic for times unknown. The way people think in this world, the way they are organised and the way they operate is highly consistent and predictable.

The world of 'Internet' though not so new, but is one that is consistently changing. 'Fail Fast' is the governing principle. MVP is a virtue. People work by collaborating with each other, relations are based on APIs, growth hacking is an obscure art and companies are built on VC money. It's okay if a company has not generated profits in an entire decade as long as they are disrupting things. The new KPIs are downloads, likes, fans, GMV, CTR and other abstruse acronyms.

Now let's look at IoT, it fundamentally consists of 5 layers: (1) Sensors (hardware) (2) Network (3) Middleware (4) System Integration (5) Applications (including Analytics). The first layer is the 'Things' developed by Qualcomms of the world and we can arguably say that is moves slow and needs years planning in advance. Upgrades are not easy to implement and so are not very often. The Network is again something that does not change much, only technologically (Unless you measure change by the availability of network in certain areas, this might be the most volatile :) ). Middleware are applications developed by the Internet Age, these are typically on the cloud, designed for interoperability and have the ability to adopt rapid change. System integration refers to integration of the middleware (or applications) with the existing applications of the client to create usefulness. Applications are a set of programs or analytics algorithms built on top of the integrated data (from middleware and other sources) to generate insights, create alerts and take actions. The applications layer is the one which has the highest flexibility to change. One other element is the after sales service and support.

The reason I've regurgitated the basic definitions is to emphasize on the fact that in the near foreseeable future no company alone can deliver an entire IoT project. It will have to be a consortium of companies in most cases. Even competing companies may come together for large or strategic projects. The reason this becomes important to observe is because the probability of IoT project succeeding depends highly on the nature of the consortium formed. The synergies between the partnering companies and the terms of the agreement among them are paramount. In any typical project the partnerships are mostly among the 'internet' companies or among the 'thing' companies. The risk there is not very high because they come together for tasks which are fundamentally same and congruous.

IoT is also about productization of these two worlds into something which is usable, manageable and yet constantly upgradable, so it also becomes important to understand who among the consortium of companies plays a leading role. So say there is a need for a company to adopt IoT, there are two ways to go about it. Either it's the Thing leading the Internet or it could be the Internet leading the Thing. How is it different? Let me entertain you. Suppose there's an electric light manufacturer who wants to think about IoT. The 'Things' perspective will lead him to put a sensor for measuring how much electricity is consumed, how it can be optimized etc. If the same problem were thought from an 'Internet' perspective, the product manager may want to understand the usage pattern of the consumer to do variety of things - repeat the same pattern when user's not at home or provide the 'right' times for marketing companies to contact the user etc. May not be the best of examples, but I hope you get the point.

This is the exact reason why I firmly believe that for any IoT project, companies should look at the whole consortium who will deliver the project, what roles do the participating companies will play, will they play roles according to their strengths and working methodology or are the roles more opportunistic. It is also important for the client to dictate who should lead the engagement. If the need is to use internet to augment the 'Things', the company responsible for the hardware part must take a lead. Whereas, if the need is to create a new business model altogether, the internet company should take a lead.