How Can We Make the IoT Work for Everyday Human Beings?

Today the IoT is an amorphous, confusing, and in some cases, downright scary combination of ideas, technologies and myths: tomorrow my fridge will pre-order all my food, my car will drive itself and my house will be programmed to control my living conditions. But the reality is much patchier than that.

Yes, there are IoT devices, networks and applications that can bring a machine-operated world much closer to reality than we might have considered possible twenty years ago. Yes, autonomous cars are driving out of test environments and onto our roads. Yes, smart homes are helping us to control our home security, heating and lighting bills and home-entertainment choices. But the gaps between the hype and the reality are huge.

Right now, the IoT is a confused, febrile swamp of technology, connectivity and creative ideas. In a bio-evolutionary analogy we are in the primordial ooze phase, and if we are going to short-cut some evolutionary dead-ends then there are some things we need to fix before we are digital/human-ready.

 

The State of Play in the Internet of Things

Essentially, the IoT is made up of three parts:

  • Devices that can monitor, report, and change their operating behaviour based on information about their environment, their status, their interaction with other devices and/or end users.
  • Networks that allow data to be communicated safely and effectively between devices, between devices and humans, between devices and analytical platforms, and between operating systems.
  • Data that are provided, collected, observed, ingested into platforms, systematized, processed, analysed and then used to develop predictive event-triggered models that can send real-time instructions to the devices to do something differently.

Right now, we have some problems…

1) We Struggle to Make Devices Act Adaptively

Most IoT devices are very, very simple – they evolved as sensors. They’re a bit like plants – give them more light and they will feed back to a central server that there’s more light. Give them more heat and they will feed back that it’s hotter. Even devices that observe AND respond to their environments; <it’s light, switch off> for street lamps or <it’s cold, switch on> for thermostats, are binary – they are either off or on. They can’t really manage <it’s light now, but the air pressure is dropping and the humidity levels are rising so it’s about to rain – stay on> or <it’s cold, but the people who live here are 50 miles away in their car and it will take them 90 minutes to arrive based on traffic and it only takes 20 minutes to heat the house up so don’t switch on for an hour>.

2) Our Networks Don’t Interoperate Well

Devices can only talk to other devices if they know what each other is called (Resource Locator) and can speak the same language (Protocol). In the Web, Tim Berners Lee solved those two problems with the URL and HTTP. The IoT doesn’t have the standards yet.

Worse, because we already do have the Internet and we aren’t starting from scratch, we’ve got multiple pre-existing systems and protocols to navigate. In the IoT, a mobile phone is an IoT device but so is my BMW car-tracker which uses a SIM. Both of those devices are communicating across the same network, however, they use different methods to do so, because the original standards required to deliver a mobile voice telephony service were different from those deemed to matter to an in-car tracking SIM. And that’s within the same technology base, from a single industry!

3) Our Data Transport is Bi-Directional at Best

Data travels from the device to a Network Operating Centre, stuff happens, and data travels back from the NOC to the device. The IoT can’t wait that long, latency is a critical operating factor. So devices need to talk to each other and act on each other without having to go back to the NOC except in cases of exception. We need a network like a spider’s web, not a network like a wheel. It’s called mesh networking, and because it can be highly complex to manage, it also needs to be self-healing. It’s a bit freaky especially if you are the owner of a big, expensive infrastructure asset, because you end up putting more and more of the value of your asset further and further away from your operating centre – where you can control it and manage the security of it.

4) Our Data-Intelligence Capabilities Lag Our Aspirations

What’s worse, even if we have adaptive devices generating data and talking to each other using common standards and sending their data to each other and to a control centre across a self-healing network, we don’t have the ability to ingest, process, understand and act on the data responsively. We don’t have consistent data schema to process the data, nor do we have enough simultaneous parsing or processing capabilities within traditional database structures to crunch and analyse the data in real-time. Furthermore, we can’t store the data adequately so that we can get access to them quickly, we fail to store data securely enough to prevent breaches, and we keep both far too much data and the wrong sorts of data because we really don’t understand what we are up to yet.

Have We Forgotten to Ask the Important Question?

When the World Wide Web was born it had a really clear value proposition – what if all the scientific research in the world could be stored in such a way that every research document could announce what it was, where it was stored, and be accessed by and linked to every other research document?

Over time the value-proposition evolved, so that instead of being the world’s reference library the Web became the world’s shop-window. The basic operating framework stayed the same, but now instead of measuring value based on the density and relevance of hyperlinks between documents, internet business measured value based on the volume and frequency of audience visits. Defining clear value-proposition metrics made it possible for participants large or small; commercial or academic; investors, innovators or users to understand how to ascribe a financial benefit to their activities, creating business cases and funds.

There is much talk about how brilliant the IoT is going to be – that once we get to the stage where it is all singing and dancing – and the heating does get turned on at the right time before we get home, and the milk does get ordered before it runs out, and our cars really do help us drive more safely or drive for us entirely, that everything will be perfect. But we aren’t going to be able to get to that utopia unless we address our digital/human-ready problems – and to do that we are going to need collaboration: on standards, on platforms, on data schema, on intelligence.

All of that activity requires funding, business cases, and clear value-proposition metrics. The IoT value proposition needs a definition, and I think it has to start with the impact on human beings.

The transformative power of the IoT is going to be huge, but it’s also going to be personal and intimate. Our lives are becoming more and more digital, and the merging of data about our surroundings with the ability to change it simply through the interaction of devices will have measurable improvements on our day-to-day existence.

Value in the IoT needs to be expressed as “the potential of using data to achieve a real-time change of state on a device to deliver a measurable improvement to the experience of a user.” We need metrics that show how many events with measurable benefit get driven by an IoT service – that move us from indexing through engaging to interacting. I want to see the next NASDAQ multi-billion dollar IPO valuation predicated on the potential of an IoT service to deliver measurable, event-driven interactions that actually improve everyday human lives.

This article was originally published on Caroline Gorski’s LinkedIn. To learn more about Caroline, connect with her on LinkedIn, follow her on Twitter or visit her blog.

Leave a Reply

Your email address will not be published. Required fields are marked *