IoT Think Tank

The Internet of Things in 2015 vs. What's Gone Before

The differences between the Interet of Things (IoT) and, say, embedded systems may seem small but they're vital -- it's those differences that will take IoT to the mainstream this year.

Last year the Internet of Things (IoT) gained a lot of momentum with a number of stories across consumer and industry markets. Almost every technology conference had an IoT element to it, and there was a proliferation of events dedicated to the subject. This year will be all the more interesting as you start to see IoT actually enter the public consciousness. Yes, many folks are aware of wearables and a few know something of the connected car, but IoT is generally an unknown. Until now it has been something for those in the technology business or for those who follow it.

2015 will be the year that many more people start to experience IoT, although they may not know it as the Internet of Things. Instead, it will just be a new device or, more likely, a new version of an old device. For many folks the benefits of IoT might not be directly visible, but hidden within the supply chain of a product or service that they use.

Aren't We Already Doing This? (aka Embedded Systems)
A lot of what will constitute IoT in 2015 will come out of the embedded systems industry, whether that be in manufacturing or in the range of other related industries such as electronics, automotive, aerospace and defense, health care, medical equipment and so on. Ultimately, it relates to companies that have a history of making devices with a software element. Some of these are small, like a home appliance, and some are much larger, such as a passenger aircraft. For those who have forged careers in this industry, the idea of actionable data coming from sensors within hardware is not new at all. And so the question these folks and others most often pose is, "What's different about IoT?"

This is a valid question, if you think about a jet engine, for example. These have always come loaded with sensors that collect data to be used by pilots or engineers. The results of that data will often lead to actions being taken, either at 30,000 feet in the air or on the ground. To try and tell the manufacturer of this equipment that there's a great new concept called IoT that allows them to collect data from sensors and take actions based on that data would be crazy. They would quite rightly respond with, "Yeah, so what?"

To answer the question requires a more specific look at what IoT actually is. As a strategist from ARM Ltd. told me: Without the Internet and Things, there is no IoT. This might seem like a silly comment, but it's the essence of the answer to the question, "What's different about IoT?" The type of solutions with which embedded systems have historically dealt are typically proprietary, point-to-point solutions. By that I mean the technology stack from the hardware to the software and the data and network in between has all been created specifically for that single-use case.

This has allowed systems companies to create solutions that serve a specific need, but they take a lot of time and money to deliver. Furthermore, they're closed systems, which means integration can be challenging, and just as timely and expensive. Interoperability has often been poor, so a lot of work goes into a scenario where one system has to work with another. Just consider complex environments like aircraft or power plants: many systems, each built for a specific purpose, having to integrate into a single environment and work together. The outcomes have often been costly and not ideal.

Take, for example, a military vehicle that has multiple GPS antennae attached to it. This is because different systems inside the vehicle require GPS, but each system is proprietary and built as stand-alone. They integrate into the vehicle but not with one another. Such is the level of independence that each system can only be operated from a specific seat within the vehicle. Should operators need to change systems, they have to get out of the vehicle to change seats. Imagine that in a battle field scenario! Alas, this example is completely true.

3 Key IoT Features: Low Cost, Standards and Availability
IoT changes this landscape by altering the fundamental architecture. The hardware can now be built from or use commercially available, generic components. These are far lower cost, easily accessible and much more interoperable than custom-designed and -made solutions. Components are already being made by manufacturers (such as ARM and Intel Corp.) and being used in commercial scenarios.

The cost is so low that individuals and small businesses can afford them, thus creating their own solutions. Where the components are already in the target environment, it's incredibly easy to plug them into a solution. This is because they're far more open than traditional proprietary hardware.

At the heart of many of these component systems is a circuit board that supports many common OSes. These include variants of Linux, including Android and Windows. Some are supporting container technology like Docker that allows applications to be easily installed and replaced over the cloud.

An easy example is to consider is the tablet as a low cost, widely available device. A manufacturer of medical equipment is using tablets to monitor patients from within their homes. The devices are far cheaper than building a proprietary piece, and they use open standards for writing code, communication and data.

The development skills to write an app that will run on an Android or iOS tablet are widely available, the technology is free to use and there are many open standards such as HTTP and RESTful Web services with which to work. Issues such as security can be managed by third-party software that has been created for the devices because of their ubiquity. In other words, the maker of the new system doesn't have to reinvent the wheel, but can benefit from others who have worked on the same platforms, of which there are many.

In addition, the device can be used for other purposes -- both by the patient and by the solution provider (for other services), and even by other solution providers. The tablet is an easy example, but individual components such as sensors can be used similarly to create hardware-based solutions quickly and for a low cost.

"Things" are just one part of the solution, the other is the Internet. The Internet provides an open, standards-based network over which these devices can communicate with one another or other systems. Whereas previously a provider would need to provision its own communications infrastructure or attempt to use a complex, proprietary third party, the Internet makes it fast, easy and inexpensive. As with the tablet example, the technology is open and familiar to many developers and there are existing third-party solutions to address concerns such as security (HTTPS, SSL and so on).

In the tablet example, the device is inherently Internet connected, so the manufacturer doesn't have to worry about connectivity. Another example of the benefit of the Internet with respect to IoT is the well-publicized smart thermostat. Whether it be the Google Nest or the Honeywell Lyric, they both make use of Internet connectivity via Wi-Fi. Imagine trying to achieve the same capabilities without a freely available and common standard network? Instead, these devices would probably have to be hard-wired into telephone cables and use telephony technology to communicate. The added complexity and expense to the user and the provider would probably render the concept dead.

The Internet also provides easy use of other systems connected to it. Cloud services are popular in many IoT scenarios -- this isn't just manufacturers writing code in the cloud, but being able to work more quickly and inexpensively by making use of existing cloud services. 

Let's not pretend that this makes everything super easy -- there are still challenges. Many of the new circuit boards require a constant connection to main power or regular battery charging. If a device needs to run for long periods without either of these, then something far less powerful is required. This changes the nature of the technology from a software perspective. Even so, access to such technology is still easy and cheap and programming for it isn't that difficult. Mass production of new hardware still requires certain skills and processes that add time and cost to a project. Again, even custom hardware can use the same standards and low cost elements.

Keep the Old, Even with the New
Returning to the original question, hopefully it's clear why and how IoT is different from the instrumentation and telematics systems of the past. That said, IoT is not a reason to rip and replace without thought. In many cases, existing systems work well just the way they are. Removing all the sensors and communications infrastructure and rewriting code is probably not beneficial, nor is it cost effective. Instead, organizations should look to new use cases.

Taking the jet engine example, there's no need to replace the instrumentation on that equipment. However, adding cheap sensors that connect via Internet to a cloud service that allows the location of the engines to be tracked at all times could have new benefits. Perhaps it can help with stock management or a reduction in parts lost, just as the medical devices company found a new service rather than trying to rebuild existing products, like heart monitors.

Moving forward with IoT, for traditional systems organizations -- with their wealth of experience in concepts such as instrumentation and telematics -- it should help them to embrace the world of IoT and not change what they've done, but enable new innovation. At the same time, IoT opens a world long closed to far more players. Already we're seeing the overwhelming majority of IoT companies being less than 50 employees.

Low cost, availability, standards and openness have historically been critical to the success of any new technology movement. If anything, it's in these ways that IoT isn't different from what has gone before. What we can tell from previous examples, however (consider the PC or Internet), is that it will change the world -- and that really starts in 2015.