This article is brought to you by ComputerworldUK in association with Intel IT Center
The Internet of Things (IoT) has the potential to drive fundamental economic and social change. However, there are serious obstacles to be overcome if we are to gain the full benefits of what could be the building blocks of a new technological revolution.
These obstacles are simple to list and hard to get a grip on, and include data volumes, data storage and management, data security and the difficulty of getting actionable real-time insights from the data flows created by IoT infrastructure.
Dealing with the analytics, infrastructure and security issues around IoT requires significant involvement and investment by both enterprise IT and vendor organisations but it will be worthwhile because of the staggering potential that IoT holds.
Spelling out the prospects, Mike Gualtieri, of Forrester Research, warns, “Don’t make the mistake of thinking that the IoT is just about industrial use cases such as manufacturing or logistics. Perhaps the most interesting Internet-connected device in the age of the customer is the good ol’ smart phone. The smart phone is laden with sensors such as GPS, accelerometer, audio, images, position, and even the touch-screen and buttons.”
IoT sensors, Gualtieri says, “emit live data such as motion, temperature, voltage, pressure, audio, video, position, and the list goes on and on.”
Analyst firm Gartner meanwhile, estimates that there will be 4.9 billion connected ‘things’ this year, rising to 25 billion by 2025.
If we are still in the early days of the IoT revolution, there are already some important and some less obvious applications of the technology. Smart power grids are an example of how IoT could lead to unparalleled levels of sustainability. Intelligent energy systems could monitor usage in real time, and analyse the data to find trends that will help suppliers to optimise energy usage according to demand.
There are smart grid pilots all over the world, but a particularly interesting one is in Austin, Texas. There researchers are carrying out a comprehensive consumer-focused smart grid project. It’s called Pecan Street, and it involves hundreds of homes, schools, businesses and one of the largest supercomputers in the world, as well as the largest concentration of electric cars found in the world.
The IoT infrastructure uses Intel technology for many of its components including the sensors in the buildings and the supercomputer at the Texas Advanced Computing Centre (TACC) at the University of Texas.
The sensor technology measures energy usage in the buildings and then streams that data to TACC. There, Intel servers analyse huge amounts of data, charting usage patterns, visualising data, and helping researchers to discover how they could implement advanced energy management systems.
The Smart power grid is one complex and impressive implementation of IoT. A more bizarre, but equally innovative example is the way the IoT is revolutionising coffee machines.
Costa Coffee’s digital barista is a vending machine that gathers information from people as they buy their coffee, so it can make the right type and combination of products available in a particular location. It analyses the purchaser, working out their individual details (gender, age etc) and preferences (coffee type, aromas), to show them targeted content on screen and even attempting to predict the coffee they’re likely to order.
To do this, it uses the Intel Audience Impression Metrics Suite, a software platform that generates real-time data analytics. Costa is planning to roll out 2500 digital barrister machines in the UK and Europe.
So IoT is innovating a large number of industries, bringing new customer experiences and the sort of service integration that companies could only dream of in the past. The problem is that creates a number of new challenges that inevitably – and quite rightly - fall on the enterprise IT team.
Forrester Research analyst Frank E. Gillett say, “CIOs will end up operating the infrastructure of the connected world, just as they ended up owning PCs, websites, and smart phones, all of which started as do-it-yourself efforts by the business.
“As they integrate IoT into the business technology agenda, CIOs will face five categories of challenges spread across the three basic domains of IoT infrastructure. To prepare, CIOs need to coordinate the work of application development, architecture, infrastructure and operations, and line-of-business teams.”
Handling the huge volumes of data that will be created by IoT infrastructures is an obvious problem that needs addressing.
Analytics software is essential and big data analytics platforms such as Hadoop have already proved their worth – storing and processing large volumes of data across a linearly-scalable cluster. However, streaming analytics platforms will also be needed going forward, which can respond in real time in order to glean immediate, actionable insights from streaming data.
Analytics are the means through which enterprises can create new efficiencies and opportunities, but even before they get there, there are other more fundamental issues that will need to be investigated first.
Storage is one. Is it desirable to store IoT data in conventional file storage systems or data warehouses? Sensor data, geographic location data, and granular customer behaviour information may call for new types of agile storage infrastructures.
Secondly, there is a need to have appropriate network bandwidth and other infrastructure in place to deal with data that may be arriving in spikes (for example if the system is measuring something unpredictable such as car traffic volumes or weather patterns) or in high velocity continuous streams.
IoT infrastructure may also need to deal with multiple data formats - more so than existing IT platforms such as ERP systems or content management repositories. For example, how do you handle the fact that there is no standard data format for IoT device data?
Fourthly, enterprises and technology suppliers will need to build platforms for time-sensitive insights that can handle real-time sophisticated data analysis. Again, a new level of agile, responsive infrastructure needs to be created.
There are developer kits available to help develop and deploy IoT programmes, and this will help. For example, Intel has created a line of intelligent gateways that collect sensor data at the network edge, then act as a filter to analyse and ‘normalise’ the data so it can be shared through the network and the cloud.
The gateways can connect together new and legacy devices and integrate technologies for networking, embedded control, enterprise-grade security, and manageability. They been designed to help enterprises quickly develop, prototype, and deploy their own intelligent gateways.
Also available are reference models and kits, such as Intel’s IoT Platform, that gives enterprises IoT components that have been pre-tested and assembled with factors in mind such as interoperability, scalability, security and management.
In addition, other vendors are working together to create the next generation of technology building blocks that will transform the world. Are you ready to plug and play?