IoT Think Tank
Debunking the Top 3 Internet of Things (IoT) Development Myths
There are a lot of misconceptions around the Internet of Things and what it means for developers. Creative Intellect Analyst Clive Howard cuts through the hype and tells you what you actually need to know about this growing area.
- By Clive Howard
Developers and IT departments are just starting to get to grips with mobile, and now focus is rapidly moving to Internet of Things (IoT). No doubt you've seen a relentless stream of IoT news lately, from connected cars and toothbrushes to home thermostats and parcel-delivery drones, not to mention the bourgeoning collection of wearable devices such as the recently announced Apple Watch. All of the big names in computing -- Apple, IBM, Microsoft, Intel, Cisco, SAP, Oracle and so many others -- have been jumping on the band wagon with offerings from cloud services to circuit boards.
With so much noise it is difficult to get a clear signal as to what IoT means for most developers ,let alone what it really means for different industries and markets.
Make no mistake, IoT offers potential and opportunity with tangible benefits and returns for a broad spectrum of organizations and use cases, both right now and for the foreseeable future. However, with all the overhyped value-add that has been attributed to IoT, it will be crucial to understand: exactly what is required, what can be delivered or achieved, and the different points that will allow for meaningful engagement and sustainable returns.
The concept of IoT is not new; we've been on this journey for a while. The move from mainframes to IP networks inside organisations is an example of an early, if limited, IoT for many companies. Telcos have also had their own early IoT environments as they transitioned from analogue to digital over a decade ago. In many respects, IoT is the logical progression that comes from the convergence of three distinct streams: the Internet, software and physical devices. It's today's manifestation of software innovation within hardware (mechatronic) systems underpinned by innovations in connectivity, from telematics to the Internet. One might even go as far to venture that IoT is the umbrella term for ubiquitous connectivity.
And the data generated as a result of Internet-connected things will place analytics at the heart of IoT.
Over the course of the coming months, this column will see different members of the Creative Intellect analyst team address the supply chain of IoT, the supporting technologies and applications. Our goal: To get behind the hype in order to expose the realities that will enable practical and sustainable engagement. We'll challenge perceptions and provide a voice for pragmatic but creative thinking.
Let's start by addressing the three most common misconceptions around IoT.
Myth 1: From Cars to Wearables, It's About the Consumer
The IoT stories that grab the most headlines are, unsurprisingly, the sexy ones. Cars that will drive themselves or Google Glass-type wearables will generate buzz. These kinds of stories have been around for a while, starting a number of years ago with internet-connected fridge/ freezers. While these are all part of the IoT collective, the genre itself is much more broad. More importantly, it will enable new capabilities that will have far greater impact on the world than your car being able to alert you to the upcoming Starbucks. In fact, you don't have to look far to see the burgeoning landscape for IoT-- from building management services and home monitoring systems, through to the control and monitoring of equipment and sensors. Many industries and markets offer ripe potential for IoT-based applications.
One such example is the use of sensors within agricultural land, which optimize the use of water to improve crop harvests and reduce water usage. When you consider ongoing food and water shortages, the impact of this is far more profound than my watch telling me if I've dropped a pound or two. A further example is construction cranes that use weather data to make sure that they don't crash into each other due to high winds. The result has already proven to save lives and reduce costs. There will of course be consumer applications of IoT that will bring about changes in behaviour and interaction, but it will be the industrial ones that will most likely have the biggest impact on the world at large.
Myth 2: Unfettered Access for All Software Developers
The buzz around IoT is leading many to believe that the opening up of hardware platforms to software developers will present a scenario similar to when Apple opened up the iPhone. This is what Steve Jobs once referred to as the modern "gold rush," where developers could make vast amounts of money by building the right app. Again, this comes back to connected cars and wearables which have companies encouraging developers to build apps.
While there will be opportunities for the wider developer community, in most cases the platforms of Internet-enabled hardware products will not be opened to everyone. The use of technologies associated with IoT such as cloud services, MQTT (a lightweight publish and subscribe messaging protocol used for machine to machine connectivity over the Internet), the better use of existing protocols as well as some new interoperability standards will undoubtedly help to drive greater innovation. There will be the potential for wider developer engagement; however, the depth of interaction and manipulation at the core platform level will be, as it always has been, trust-based.
What will be important is in the way that physical device providers open up their product platforms to easily enable third party developers to interact with it but also to monetise their output.
Universally accepted standards, protocols or platform environments, along with UX considerations and a solid program for partner/ecosystem support, will allow for more rapid innovation to come into the traditional embedded systems market through external sources. However, as pointed out earlier, deeper levels of product innovation will still only be open to a smaller group of trusted development partners. Such trusted partners will be able to work more directly with the platform's core facilities. A wider group of developers will then be able to access and manipulate data or platform services through controlled and managed interface as is done today within many platform technologies. It is a tried and trusted pattern that works.
At the outer edges of engagement, the vast majority of developers are more likely to use the variety and volume of data that will become available to drive new use cases and experiences. You can see how this is happening today at smartthings.com.
Crucially, for most companies, their IoT platform will be bounded by their business. For example railway operators and infrastructure providers are likely to see their IoT platform as being track, signalling, rolling stock, message boards and ticket offices. Access to these systems will be limited to a few partners who will then extract data such as train locations, scheduling and ticket prices to pass on to customers.
From another perspective, Ford may want developers to build apps for their cars infotainment systems, or leverage data that the company can easily and securely expose from the vehicle's internal systems, it will only allow trusted and highly certified partners intimate access to lower level system control units (internet enabled or not). An example of this will be their dealer and service networks who will use the data to do real-time monitoring of a vehicle's health. Where a problem occurs, they can then ask the owner to bring the car in. This system will almost certainly then extend to support fleet managers who will want to track driving standards, behaviour and even use it for finding lost or stolen vehicles.
Ultimately, it is not just car manufactures but all transport systems that will look for applications to be written by thirds parties to engage their customers that will find value in IoT. But this will not translate to unfettered access to all areas of the codebase.
So while IoT will draw many software developers into the systems space it is not going to be opening it up to everyone as mobile did.
That said, there will be other opportunities for software developers as hardware becomes more accessible. We're seeing companies like littleBits and Bare Conductive offering innovative ways to very easily create new devices.
The barriers to entry into the hardware business will come down and software developers will be able to get into the IoT market. A recent example was of a mobile app company that attached Wi-Fi cameras and tablets to delivery vehicles. Drivers were then able to view their cargo being loaded and unloaded via the tablet in the cab. The result allowed drivers to collect and deliver without getting out of their vehicle. In this case their cargo happened to produce a large amount of dust that over time could be harmful to the driver's health. This shows how software companies are already able to easily work hardware into their solutions and the outcomes have value beyond reducing costs.
Myth 3: It's About a Gazillion Terabytes of Data
There's no doubt that IoT will mean a lot of new data being collected (so called Big Data), but volume of data shouldn't be the ambition. The value of data within IoT should be the actionable outcomes that it enables, and not simply collecting as much data as possible. Businesses will need to think about what data they need to derive value and then collect just that data.
An example might be monitoring wear and tear on a particular piece of heavy equipment. The value might be in predicting potential failure and so enable early maintenance which will save downtime and cost. To achieve this, certain data points will need to be collected and analysed and it is those specific data points that the solution should focus on. It will be tempting to gather every piece of data possible but this will increase data transfer, storage and back-up costs and the business will have to justify whether this makes sense. In most cases it probably will not.
From a technical implementation standpoint, IoT is about being fast and lightweight. If a device is sending data over the network regularly then protocols such as MQTT will be highly valuable as they reduce the size of the data sent (size = speed and cost). It is also worth remembering that a constant connection with the device may not be possible and so data may need to be temporarily stored on the device and then synced over the network when connectivity resumes. For reasons such as these the ambition should be to minimise the data collected. Being targeted and focused on the outcome is vital. Big Data and Analytics is about empowering systems and users to deliver value added outcomes, data analysis should never be the purpose in itself.
The analytics is more important than the quantity of data, purely because the more information that is coming in the harder it will be to identify what is essential and what is just noise. Today this is an area that needs further work from the vendor community, although the likes of IBM and SAP have already begun to deliver their solutions.
An IoT truth: Smart Devices Need Smart Businesses
There is no doubt that the future will see us surrounded by (even wearing) devices that will be smarter and more connected than those we have today. This will have implications for consumers and industry and for the underlying networks and infrastructure. It will impact professionals within IT and also within the embedded systems arena. The businesses that produce these smart devices will need to be smart about how they embrace these new possibilities. For developers it will mean new technologies, services, processes and tools. But most importantly it will mean understanding how to apply these in a way that will really make a difference. They too will need to see through the hype and noise to new opportunities.
Going forward there are questions to be asked concerning the security of IoT based products and systems. This is something we will cover in later column discussions along with the role and strategies for analytics.