IoT — Getting it Right
Why now, and how to avoid making a bad IoT product
SafetyCulture is in the process of researching and evaluating a wide variety of IoT hardware and potential applications, to enable businesses all over the world to improve their safety and quality controls in their day-to-day operations.
In this article, we will share many of those learnings by exploring recent trends in the technical development underpinning IoT. This will include explaining terminology, key principles, philosophies and our predictions in the IoT space.
In future articles, we will evaluate IoT hardware including development chipsets, industrial IoT and consumer products, and dive into the network infrastructure that is shaping up to underpin an explosion in IoT products and services. We will also reveal some of our key learnings and plans to scale in B2B IoT markets.
What actually is IoT?
The word is thrown around so much in tech circles that it’s easy to forget it has any meaning at all, beyond describing the next generation of tech development.
Internet-of-Things as a phrase is vague, and to assume that “things” connected to the internet is a new phenomenon would entail a lack of appreciation of all efforts in miniaturising wireless technologies that came before it.
Nevertheless, it has become apparent that the term “IoT” is a powerful tool to describe the permeation of smart devices, sensors and communications technologies.
To understand what an IoT thing actually is, we should form an opinion on what it is not.
There is no precise definition of exactly what an IoT device is, but broadly it includes any embedded system with internet connectivity. However, for the meaning of IoT to make sense in a modern context, the definition should exclude already universal personal consumer products such as smartphones, tablets and laptops.
The reason for this is simple: the purposes of IoT include gathering information from the environment, objects and people, and communicating that data to provide meaningful insights on those entities, thereby allowing action on that information physically or virtually, at immense scale.
If this data is going to be truly valuable, enabling us to monitor sites and equipment, track movement of assets, and improve our well-being, there will need to be many more data sources than exist today. Instead of 1-to-1, we should expect the ratio of internet-connected devices to people will be 10x or 100x.
That means potentially hundreds of billions of IoT devices in circulation within the next few years.
A few conditions need to exist for this to happen. Fundamentally, adoption of IoT is and always will be driven by a simple cost-benefit calculus. We need IoT to be cheap, suited to purpose, appealing, and easy to use before we can expect the rapid uptake predicted a few years ago.
Why are we talking about IoT now?
While rhetoric almost always circulates before any meaningful technical development, we can now say with confidence that the development is catching up. The pre-conditions of cost and usability are being met, to the point that previously foreseen applications of IoT technology are now becoming practical.
Recent developments affecting IoT can be categorised into four key areas:
- Embedded communications chipsets;
- Network infrastructure and standards;
- Cloud infrastructure; and
- Power storage and optimisation.
Together, these have removed the pain of having devices that are bulky and expensive, difficult to connect, perilous to maintain and run out of battery faster than you can find the matching charger.
While we will focus on the first two in the above list, the importance of the others cannot be discounted.
There has also been a marked increase in awareness of the capabilities of IoT and the value those capabilities can unlock, particularly in a business context.
Instead of 1-to-1, we should expect the ratio of internet-connected devices to people will be 10x or 100x.
This has been partly driven by a number of iconic software and hardware players (Amazon, Google, Microsoft, GE to name a few) introducing new products and services under the banner of IoT. In corporate circles, most new business analysts, digital transformation managers and c-suite executives may fear being behind the times simply by not mentioning IoT enough.
However, the vast majority of the focus from the big players still remains in the cloud space, as few dare to venture into the costly and ever-changing playing field of IoT hardware.
So the question remains: who is here to truly connect our world with IoT?
Edge-compute, blockchain and other silly obsessions
Let’s be clear on the problems that IoT is solving.
Although some of the hype is now backed by technical capability, there is quite frankly a lot of useless discussion in the name of IoT around ideas that have nothing to do with IoT, or are so premised on IoT being universally widespread today that they could best be described as delusional.
IoT solves first and foremost the connectivity problem. The vast majority of potential physical data sources today remain untapped, because there is nothing yet to connect them in an easy and cost-effective manner.
If we could simply attach a power-hungry, fragile and expensive compute-module to everything we want to understand, wouldn’t we have done it already?
For now, the talk of adding substantial edge-compute capabilities to IoT devices is premature at best, and at worst, misses the point entirely (with the notable exception of critical real-time applications such as driverless cars) by returning us back to the original problem of having expensive and power-hungry devices.
Blockchain in an IoT device may provide a fancy way to track your asset, but it won’t stop someone from stealing it.
In many cases, what could be solved by “smart” IoT devices would be much better solved with scalable cloud solutions paired with simpler hardware. This has the advantages of being lower cost, more powerful, versatile and infinitely more reliable.
The technology that makes IoT practical is new, and as such IoT is still in its infancy.
Our focus should be on making IoT as low-cost and efficient at communication as possible, enabling more businesses and consumers to get their data where it needs to go, at which point we will have all the power to understand it. This means devices that are smaller, cheaper, and last longer, to fulfil a simple purpose, such as measuring a temperature, monitoring power consumption or tracking movement, as reliably as possible.
How to connect your fridge to the internet
So what comes next for IoT?
We live in a world where the vast majority of goods and infrastructure remain disconnected from the internet. So how do we achieve an IoT nirvana where we have wireless monitoring for everything from office buildings, to rubbish bins, and even your package from amazon?
By understanding technical capability and economics, we can predict that the deployment of IoT will likely go through three separate waves:
- Retrofitting products with devices connected to local networks;
- Retrofitting products with devices that connect directly to the internet;
- IoT embedded in the majority of new products.
The reason for the above progression is two-fold.
In the short-term, it will be cheaper to fill the lag in replacement of old assets and products with retrofitted sensors. In future, we can expect that manufacturers will include low-cost sensors embedded in the construction of commercial products as well as larger structures, as a key value-add of any goods that they will certainly need to keep themselves in business.
The other reason comes down to the current state of IoT networks, as well as the chipsets able to use them.
Low-energy IoT networks such as Cat-M1 and NB-IoT are on the cusp of being ready for wide-spread utilisation. But due to limited global coverage, lack of tel-co commercialisation readiness, and the immaturity of systems able to use those networks, the next-generation of IoT capability just isn’t ready yet.
This will very likely change in 2019 & 2020 as the IoT arms-race gathers pace.
In the meantime, most IoT deployments will likely involve devices that have hard-wired power that is painful to setup so they can use Wi-Fi or 3G/4G, or require local network infrastructure such as LoraWAN.
Think customer!
All of the above sounds quite painful to consumers and businesses alike who want to adopt IoT.
This is true. Whilst user-experience and customer-centric design have been core principles of software companies for some time, there is not much common understanding today about what this means for the next generation of hardware.
The variety of embedded devices that have been through the rigour of consumer-based testing en masse is quite limited: PCs, printers, Wi-Fi routers, and smart phones & watches. It usually takes many years for device manufacturers to iterate on designs and find the winning formula, and the gradual education of a tech-savvy consumer-base to be able to use them.
For those developing IoT products, it would be unwise to wait another decade to make IoT foolproof and useful enough for the majority of businesses and consumers. But until that happens, we should expect IoT to experience slow growth.
Fortunately, thanks to all the low-cost electronics and cloud infrastructure that has been developed over the past decade, we now have the ability to build products that involve hardware in agile.
Interested in solving IoT problems that truly impacts people’s lives?
We’re hiring.
