Real-time data analytics | Director


Thanks to real-time data analytics could the information revolution be entering its most disruptive phase yet? The ability of businesses to receive operational data in real time, and react accordingly, marks “the beginning of a new economy”, say experts. Here’s what the most advanced companies are doing, and how you could cash in…

It’s a dry, unwieldy phrase: one likely to have linguistic pedants asking sarcastically, “What, as opposed to ‘fake time’?” But ‘real-time’ technology – computer systems that process and relay data to users with negligible to no delay, allowing for instantaneous remote response – is fast putting a serious wind in the sails of disruptive phenomena such as open data, big data and the Internet of Things.

“Real-time analytics banish the latency between data and action, enabling today’s decision-maker to enjoy a broader, more synchronised view of their operating landscape than ever before,” says Dr Christoforos Anagnostopoulos, chief data scientist and co-founder of machine intelligence company Mentat Innovations, as well as a lecturer in statistics at Imperial College London. “It equips professionals with early warnings that help mitigate risk and fast-paced insights that generate short-term opportunities for growth.”

The world of real-time data analytics (RTDA) is shaping up to be one in which terabytes of zeros and ones bounce to and from every nook and cranny of manufacturing centres, ensuring that industrial equipment never fails and supply chains are met with up-to-the-second accuracy. In this ultra-efficient tech-topia, retail intelligence has moved on from greeting supermarket shoppers with the appetite-spiking aroma of hot croissants, and now entails sucking reams of information about customers’ spending habits from their smartphones and beaming back carefully targeted discount offers as well as making spontaneous pricing and product information decisions.

Train tracks are dotted with sensors that predict mechanical faults long before they happen and communicate them instantly via ‘telemetry’, prompting pitstop-rapid maintenance. If your business involves company cars, expect your fleet administrator to soon be receiving and acting on a flurry of warnings from hundreds of sensors fitted to each vehicle, alerting them to every soon-to-be-worn brake pad, overheating alternator or corroded valve stem seal. Smartphones double up as GPs, lawyers and financial advisers, cash only changes hands in backstreet gambling dens and the national grid anticipates demand surges and generates instant bespoke personalised tariffs incentivising off-peak consumption in response.

Already, thanks to the presence of real-time data analytics in our working lives, much that would once have seemed unthinkable has become routine: not least in the public relations sector. In the social media age, with an average of about 6,000 tweets posted per second, fulfilling that industry’s core duty – to harvest, digest and act upon perceptions circulating in the wider world which may affect an organisation’s reputation and welfare – could easily have been like tasking a lone lifeguard to watch over a storm-lashed Pacific. Instead, social media ‘listening tools’ such as Brandwatch, Hootsuite and Crimson Hexagon have stepped into the breach, and now crunch the reams of relevant data out there before providing media professionals with up-to-the-second, user-friendly stats, charts and ‘peer-index influence graphs’.

Using the same platforms, marketing professionals can now keep an eye on the whimsical rips and tides that make up the collective consumer psyche, and adapt and target content accordingly (“We used to carpet-bomb – now we use guided missiles,” as Philip Trippenbach, digital strategist at Edelman, wryly observes). But it’s not just PR and marketing professionals who have cottoned on. Forward-thinking leaders in various sectors are already using real-time data analytics to monitor key performance indicators, sales, site visits, clicks, conversions, cost-effectiveness and logistics at a far higher frequency than ever before. Energy consumption alone can be measured at 15-second intervals thanks to the emergence of smart metering.

Some wily firms are coming up with real-time innovations specific to their raison d’etre – outdoor advertising firm JCDecaux now gets its bill posters to photograph proof of posting in situ and delivers the photo, along with UPS tech proving location, date and time, directly to clients – while others, such as Fasteroute, OpenSensors and Laundrapp (see box, overleaf), have harnessed real-time data analytics as a commodity in itself. In other words, organisations are harnessing the awesome power of real time. And there’s one rarefied realm that is at least half a lap ahead of the chasing pack, one that should be watched as a bellwether for future opportunities.

Track record

Real-time data analyticsWhat do you get when you combine huge budgets, world-leading expertise and a sport in which every millisecond counts? Formula One has always had technology at its core, but now broadband speed is as integral to success as the rate at which those beefy tyres cover the tarmac.

An F1 car has got 200 sensors on it, monitoring everything from tyre pressure to drag, via aerodynamics, downforce, weather and the locations of other cars. A thousand or so channels continuously relay information, milliseconds after it is generated, to drivers, engineers, mechanics and strategists. No wonder, for the last three years, F1 top brass have worked with Tata Communications to bolster trackside connectivity from 100 megabits to at least one gigabit per second at all the world’s major tracks.

“All the information that comes off a car comes into a data store. We then pull it out and convert it into focused visualisations for different people in the team, all in real time,” explains Colin Burrell, a Williams client executive at Avanade – the London-based technology solutions company tasked with improving Williams’ ninth-place 2013 finish by sharpening its IT infrastructures and processes. Burrell operates in a notoriously secretive industry, but does expand on some of the remarkable ways in which RTDA is shaving split seconds off Williams’ performances.

“There are feeds giving us tyre pressures, temperatures, performance and so on and we can build a graphical representation of how the car is performing,” he says. “We can then overlay that with how competitors are doing – it’s in the public domain – then trace the performances in real time. If we deviate from that forecast, we can visualise that and respond to it. Before, they’d have taken mathematical guesses at how well the car would perform, but if it didn’t perform there wouldn’t have been much they could do. They now get an easy-to-digest representation of everything they need on the pit wall.”

It’s impossible to overstate the extent to which real-time data analytics has overhauled the sport, according to Graeme Hackland who, as Williams’ IT director, works closely with Burrell. “When I arrived last January,” he says, “I asked the chief strategist here, ‘What do you do at the track as the cars are going round?’ He was spending 70 per cent of every lap manipulating data from different sources, putting it onto spread sheets. He’d then have the remaining 30 per cent of the lap to analyse it before the next lap started. Now, he spends no more than 10 per cent of that time manipulating information.” The accumulated data influences every decision – whether to tweak the driver’s line, visit the pit or replace parts, with 3D-printed components now even being created on the spot.

Hackland also points out the extra benefits of real-time data analytics now that F1 regulations limit the number of engineers who can be located on site at any of the 20 circuits, dotted as far afield as China, the US, Australia, Singapore and Monaco. “Our chief technical officer can work from home,” he says. “On race weekends, he can stand in his kitchen and get real-time information from a pit wall anywhere in the world, just as if he were at the track.”

Implications for SMEs

Of course, none of the above would be possible without lightning-speed connectivity. AT&T – the telecommunications outfit behind Red Bull’s trackside endeavours – provides a connection that has a maximum delay of 300 milliseconds: literally the blink of an eye. And it’s not just improvements in connectivity hogging centre stage: robust queuing systems, crowdsourced data proliferation, scalable cloud processing infrastructure, better portability and the rise of machine learning are all factors set to make RTDA omnipresent in our personal and business lives.

How can modestly sized enterprises, as well as their larger rivals, seize the initiative and start reaping the rewards? According to Chris Nott, executive IT specialist at IBM, prime opportunities are there for the taking for retailers, telecoms companies and finance firms. “The desire of such organisations to tailor their engagement with customers is one driver for using real-time data analytics. Success depends on being able to intervene in real time – it’s no use coming up with a promotion to encourage someone to buy an item in a shop after they’ve walked out, nor to deal with a customer service issue after a disgruntled customer has taken their business elsewhere.”

Combating fraud is another major draw, says Nott. “There are now applications to counter attempted payment fraud before transactions are completed,” he says. “Speed is key to beating electronic payment fraud because patterns must be detected before the transaction is completed.”

Predictive maintenance is another area where real-time data analytics has vast cost-cutting potential. “Let’s assume your dishwasher breaks down,” says Anagnostopoulos. “It’s likely that it was vibrating oddly but imperceptibly for a couple of weeks in advance of that failure. If that had been detected in time, a loose screw
could have been tightened – at a cost of £30 for an engineer visit, as opposed to £300 for a new dishwasher. Now scale this up to a car, a bus, a train, a wind turbine, a gas pipe or a nuclear reactor, and you get the picture.”

So are there any downsides? It’s not difficult to catch a whiff of Orwellian dystopia about real-time data analytics – investors seeing real-time financial figures and acting on damning snapshots of reality; health insurers making us wear data-sprouting implants to determine the cost of our next premium; litigious mayhem over ever-more complex privacy and IP issues. Those who heard Stephen Hawking’s and Elon Musk’s warnings about artificial intelligence may also be wary of RTDA’s potential to further its developments. “Machine learning’s ‘speciality’ is to give machines the same cognitive abilities as humans,” says Anagnostopoulos. “It’s no exaggeration to say this represents the next industrial revolution.”

William Higham, founder of agency Next Big Thing, however, insists this is not a phenomenon to be feared. “What we’re talking about is simply a more efficient, technology-based variant of what most SME owners do anyway,” he says. “Anybody who runs their own shop will notice customer behaviour as it unfolds in front of them, and will change things – stock, layout and so on – accordingly. A lot of 1960s fashions were created because designers such as Mary Quant and John Stephen saw people in their shops trying to pull a jacket down or lift a hem up. All real-time data does is make this happen more easily, efficiently and on a much bigger scale.”

But what about privacy? Our gradual acceptance of sharing personal data brings the phrase ‘boiling a frog’ to mind. “It’ll get to the point where we take it for granted,” says Higham. “There will always be a place for individuals to analyse that data – I don’t think we’ll ever be beholden to it. Whether that is within the organisation or an outside consultant, somebody needs to make decisions around the data streaming in. The algorithms that help make decisions will improve but decisions will still need to be made by humans. Any issues of trust will benefit SMEs more than big corporations, because people are more likely to trust a smaller entity with their information.”

What’s clear is that real-time data analytics needs leaders to respond – and with as close to zero latency as possible. “Real-time analytics marks the beginning of a new economy,” concludes Anagnostopoulos. “It’s one in which ubiquitous sensing and machine intelligence are seamlessly integrated with the human decision-maker, offering previously unthinkable efficiencies across existing markets, and creating altogether new ones.”

Read original article here: