Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

The internet of platforms: The success of IoT depends on them

So, it feels like we have been talking about the internet of things forever. And as someone that is a guilty party, I think it is only fair to share what I am seeing and what I believe will happen next.

When Kevin Ashton used the term “internet of things” back in the year 2000, he referenced a world where things would each have an address and be a member of the internet. Considering this started in the MIT RFID lab, today it’s easier to understand how he could have felt that way. Each RFID tag was a way to identify the asset, and when you think the internet was already blowing up, the notion that future identification could be embedded in “things” and they could have their own address to connect to the internet makes perfect sense.

However, one thing that was never quite clear was how all these devices would get to the internet and interact – cue the internet of things platforms.

Battle of the platforms

Like every new technology cycle, there must be a battle to claim leadership and differentiation, from the famous ones like VHS and Betamax, to TCP/IP and Token Ring to MQTT and CoAP. Battles like the rap game exist for us in the tech industry to help us understand the differences, benefits and capabilities of the options. This learning curve is part of the curve to mass adoption and use.


Network protocol candidates

To this effect, I remember the battle of the platform event and it felt like every player with ambition in the IoT space had to show up and defend their strategy, architecture, protocols and APIs in front of an audience mainly made up of techies and competitors. A look at the criteria and you immediately notice there was a common framework to build the IoT platform.


Common IoT architecture and framework

A close examination of the battle participants and other early IoT platform vendors also reveals players that were focused on serving any person or organization that wanted to enable IoT in their business — and the offerings were generic and required a lot of further development and expertise to deliver a solution to market. That is a challenge and several analysts have pointed to this being a key factor in the slow adoption of IoT especially in the enterprise space.

The shift

This approach was challenging and several players that were focused on truly delivering IoT solutions to their portfolio or customers started reexamining their strategy. A key example was a popular garage door manufacturer that had launched a solution using a generic platform, but then made the decision to self-build in-house — and it was not alone. The deep domain enterprises started exploring the landscape and asking themselves the question – should we build or buy?

Like any infrastructure decision, the build versus buy argument usually favors the buy — especially if it is generic infrastructure. However, this is where I see the difference: IoT platforms are not generic infrastructure. The high-level components might be, but the domain components and business outcomes are what make it successful.

The difference

For simplification, let’s call the early IoT platforms “generic platforms” and let’s call new domain-powered IoT platforms  “industry platforms.” The industry platforms come to the party with several industry-specific criteria that relies on domain knowledge and experience. The platforms support pre-integration, scalability, resources, information and a set of capabilities that comply with the industry requirements such as security, regulation and APIs.

You are probably wondering if this is sustainable. I believe so. First, it is important to stop thinking about platforms as a product or solution, but as an infrastructure tool. If you agree that your business is going digital and data is going to rank up there as one of your most valuable assets, then you need a platform. Do you have to build it? Not really, but you must define and own the capabilities and ensure it matches your industry.

I have spoken to several analysts and colleagues, including my good friend Brad Nicholas at Uptake about this phenomenon, and we both agree this is a major shift in the IoT space. Below is my conclusion.


Digital giants:  I strongly believe there will still be a few players that will eventually fall into the generic platform bucket, but I will reclassify them as the digital giants. And they will not just be offering a pure-play platform, but an ecosystem – think Facebook, Google, Amazon, Tencent, Baidu, Alibaba. For folks in the enterprise, you might think this does not apply to you, but think again. These folks will be part of the ecosystem if your business is delivering a digital solution. Digital giants will bring tools and capabilities that open new markets and industries, for example, conversational AIs. I expect some companies to go out and build, but the likelihood is you will be playing catch up to the giants, so a better strategy could be to adopt their conversational AI to enhance your solution.

Recommendation: Partner or collaborate
Platform collaboration: Now, if I believe in the rise of the industry platforms, then surely there must be a rise in platform collaborations. Industries do not exist in silos and their platforms will become the glue to collaborate and even compete with competitors in their space. Take population health: you need a significant amount of data for it to start making sense, so I see EHRs, hospitals, universities and even cities working together to collect the data and even do some joint research. At the same time, I expect hospitals to use the data differently and offer services to their patients. Collaboration will bring new capabilities, new ideas, faster innovation and faster time to market.

Recommendation: Embrace it and be open
Algorithms: When you look at IoT platforms, they promise to deliver business outcomes at the end of the funnel. I expect these outcomes to be algorithms that positively alter the behavior of billions of global works. They will use behavioral, psychological, social and cognitive science to determine our next best move and action. A lot of workers today perform work in a very linear way, where their tasks are not affected by real-time events happening in their environment. With industry platforms and domain knowledge, I expect to see that change.

Recommendation: Collect, arrange, share, evolve and protect algorithms

The bottom line

There are a lot of changes still coming to this IoT space and many more generic platforms. But for mass adoption and successful large-scale deployments, we need industry platforms and domain injection into the generic platforms.

Businesses are stepping up with industry platforms, and industries are enthusiastic about the opportunities, but there is still a long road before domain-based IoT is ubiquitous.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.