Data analytics from the Internet of Things

There is a familiar ring to the notion of the “internet of things”. The term, like “big data”, has sprung into mainstream consciousness over the past two years as consumer products, such as the Nest thermostat and Fitbit activity-tracker bracelets, have introduced the concept of devices that can automatically communicate with a smartphone or laptop.

However, there is little new about the concept of machines talking to machines. Such technology has gradually been adopted by the logistics, shipping and retail sectors to track goods and automatically alert stockrooms that a product has been delivered or that a shelf is empty. Yet it is only with the advent of 4G and processors that cost tens of pennies that the internet of things (IoT) has started to gain traction.

Few doubt the IoT will soon become prevalent even if there is as yet no single view of what it actually means. Ericsson currently forecasts there will be 50 billion connected devices communicating with each other by 2020. That is ten times higher than the Organisation for Economic Co-operation and Development estimate for the number of devices currently connected. Cisco estimates that 99.4 per cent of “things” that could be connected have not yet been done so. The next big thing could truly be a trillion small things.

The power of the IoT is already becoming apparent across a number of industries. Rio Tinto is remotely controlling a number of its trucks used in West Australia which is generating $300 million in cost savings. Smart water grids are predicted to generate $24 billion worth of revenue by 2020 in Europe alone as IoT is harnessed to reduce the huge amounts of leakage. Hospitals will be able to use remote monitoring on patients with something as simple as a plaster with a processor embedded, which could also generate billions in savings.

Morgan Stanley, the US investment bank, believes that the IoT is the “next revolution in computing” that will result in “an army of tens of billions of robots that make our lives easier”.

TRANSFERRING PROCESSES

There is already evidence emerging that different processes can be transferred to different realms within the IoT. SAP is now using a Sybase system, designed to help financial services companies deal with floods of real-time data, to instead process information generated by thousands of sensors installed on bridges. Morgan Stanley cites the example of a company that owned both an elevator business and building leasing company. It was able to use data generated by the usage of lifts to predict when a tenant was winding down its lease.

The big benefits could initially be felt in the public sector, which is always under pressure to cut costs. The move towards “smart cities” may seem a long way off, but the IoT is already beginning to be implemented for some of the most time-consuming and resource-sapping jobs councils have to do. Westminster City Council has installed solar-powered bins in locations including Knightsbridge and Belgravia. The bins use telemetry and infra-red sensors to determine how full they are and to alert council workers when they need to be emptied. This has resulted in a 60 per cent reduction in bin collections. The sensors will also tell the bin when it is too full so that it will stop compressing and block any more litter being shoved in.

The next big thing could truly be a trillion small things

Ed Argar, Westminster City Council cabinet member for city management, transport and infrastructure, says: “These hi-tech new bins are just one example of how we are using new technology and innovation to find ways to improve the services we provide, make them more efficient and save money into the bargain. It’s a win-win.”

ANALYSING BIG DATA

The amount of data produced by billions of sensors embedded in cars, in homes and on people will generate a staggering amount of information across disparate systems. How that information is collated, analysed and made useful is a key hurdle, not just for the technology and telecoms sectors, but for any company of scale looking to tap into the future.

According to Martin Garner, an analyst with CCS Insight: “Connecting the devices to the internet is just the first step. Most of the value will come either from controlling the devices or from using the data they generate. Most organisations are not geared up for this yet, even if they have experimented with IoT.”

Companies will at the very least have to hire more specialists to analyse data and will need to invest in IT infrastructure to capture information. Some of those investments will be huge. “Auto manufacturers will collect terabytes of near real-time data per day from connected cars. To get commercial value from this they will need expanded data centres and departments of analysts, as well as new business processes for acting on the information they receive,” says Mr Garner.

The key danger is that irrational fears over the use of data generated by our interaction with the IoT could provide stumbling blocks to its use. Stephen Pattison, head of public affairs for semiconductor and software design company ARM, says a clear system needs to be established that gives people control over their own data. “If IoT is not seen to have empowered people then it will fail,” he says. Mr Pattison cites loyalty cards used by supermarkets as an area where trust has been established and that sort of confidence needs to be instilled across the whole IoT sector.

The vision of the connected car and the smart city may still appear fanciful, but so did the smartphone only ten years ago. Companies developing big data strategies need to have one eye on how they will tackle the IoT in the future.