When Kevin Ashton coined the term Internet of Things in 1999, prophesying interconnected smart objects running the world, he was really ahead of the times.
In 1999, large parts of US, not to mention other geographies, still used dial up Internet. There were no smartphones or tablets – in fact the Personal Digital Assistant, which were nothing more than digital diaries, were a rage. Online brokers used to charge up to $20 for a single transaction executed via Internet on a stock exchange. India had just seen 2G mobile phone network launched. World over, one way pagers were still the most reliable way to reach the IT engineer on the night application support shift. And yet, Ashton peeked into the future.
Nearing two decades from the time the term was coined; Internet of Things has variously been called Internet of Everything, Internet of Things, or Connected Devices. While we see several manifestations of the phenomenon today in our day to day life – some visible, and some hidden – we also face a start reality. The world hasn’t yet become Internet of Everything; in fact it would not be far from truth to say that we are nowhere close to connecting all objects around us in near future.
So while every digital disruption makes waves, why has Internet of Things still traveling on the upward slope of the hype cycle? The answer lies in one key characteristic of any platform concept – abundance. While all multisided platforms thrive on abundance of participation from those who deliver value and those who seek it by matching them together, too much abundance is actually confusing. If the delivery agents and the seekers cannot be matched efficiently or are unable to discover right value, the platform loses its charm and with it, the economic utility.
Today it not uncommon to find senior executives eager to reinvest themselves digitally, trying to push ideas about connecting everything from machines to consumer devices to CRM software. But ask them what they intend to do with data emanating from a shop floor in conjunction with the data from a television, and they will not have a firm view. However, in the age of FOMO – fear of missing out – most of them refuse to acknowledge the seemingly obvious possibility – the two data sets aren’t related in a way which can help the firm make any immediate strategic or economic gain. And hence it is completely alright to either not link them or not even link physical objects from all parts of their business value chain.
When Internet proliferated on desktops and later on laptops, multilayered security evolved rapidly and in step with the connectivity. From the traditional n-tier architecture to the user side vulnerability protection, a whole industry worked to ensure that gains from connectivity was not compromised by downsides of risking machine interaction. When Internet moved to mobile devices, leading hardware vendors worked to bring the same level of protection to the end users on these devices. We still have security and data breaches, but in the larger context of the enormity of Internet, these are blips. Also our security systems continue to evolve, learn and self-adjust to newer dangers of hyper-connectivity.
Now think along these lines with respect to say a connected home. Let’s say we have all devices in a home talking to each other. These devices may be manufactured by any number of vendors. Presumably a living room can have devices ready to connect with each other manufactured in multiple continents by dozens of firms, who have no idea, how the others have implemented connectivity features. A service provider can create a platform which sits on top of all devices to orchestrate their communication. But every configuration of the device talking to the platform is still bespoke. For device manufacturers, connectivity features are add-on, a new way to dazzle the user – hey look, I can switch on this fan in my living room through my phone when I am in the office! But there is very limited standardization for how you switch on a fan versus how you switch off a smart television.
Now this does not imply that device connectivity has no useful purpose. Firms like General Electric have created an industrial Internet, where their machines can be monitored and adjusted remotely because they are ‘fully visible’ hooked on to a network. A mining company operator can adjust how trucks are moving several hundred feet underground, sitting in a plus air-conditioned office. But each of these scenarios is predicated over a common theme – limiting abundance.
Successful and implementable Internet of Things is characterized by a limited set of things talking to each other for a well-defined, limited purpose, with tight boundaries around the nature of communication and data being collected from these things. Each use case result is tied to a pre-determined set of physical world outcome. And while experimentation is completely kosher, the starting point of a workable Internet of Things is a limited central theme – there is no spray and pray approach.
And in this sense, what is working well is to create an Internet of Some Things, not Internet of Everything. When the scope is limited, a firm can provide the right standardization of infrastructure and communication protocol and the right level of security ring-fencing working against any attempts to sneak in the connected devices network. There can still be platforms which aggregate dissimilar things, but each platform has a known, defined utility and boundary.
As we move from solving the infrastructure and data challenges of Internet of Things, global standards will have to evolve and will have to be implemented by truly forward looking device manufacturers. Only when connected cars from factories in Seoul and Detroit understand each other can the aspect of community or commerce evolve between objects. Until then, the focus should be on plumbing rather than marketing.
The grandness of the vision Kevin Ashton put forward will indeed be realized in future. The grandness of marketing presentations from vendors offering third order connectivity-driven use cases may however have to wait for the time being.