IoT Blog | February, 2018
Head of Business Development, Vodafone Internet of Things (IoT)
The Internet of Things (IoT) is generating vast quantities of data, but research suggests that up to 90% of this is either unused, or underused.1 As the number of devices increases, the challenge of getting the most from all that data will grow. Getting to grips with big data and artificial intelligence (AI) could help you put this data to work and get more from your investments in IoT.2
When we first started publishing the Vodafone IoT Barometer3 in 2013, adoption was just 12%. Since then adoption has more than doubled, standing at 29% in 2017—and nearly all of the organisations that have adopted IoT report that their initatives are delivering a return on investment (ROI). But are they missing something? If your organisation isn’t leveraging the vast amounts of data generated by IoT, you’re not harnessing its full potential.
The traditional approach to data analysis was to come up with a hypothesis and then test it. If you made ice-cream you might suppose that there was a link between sales and the weather. You could run some queries and find out if you were right or wrong. You might come up with a more interesting, and less obvious, hypothesis, like sales being connected to sporting events or the most popular flavour following a particular TV show.
But there’s a limitation to this approach—our human mind. No matter how imaginative we are and how quickly we can come up with hypotheses, we simply could never match what a computer can do. For example, how many attempts do you think it would have taken to come up with the hypothesis that people who fill out their loan application in all caps are more likely to default? Zest Finance found just that.4
Leveraging big data with AI
Many IoT applications generate lots of data, but this just gives us more data, not greater insight. This is where AI and machine learning comes in. These technologies can identify patterns in data that we could never hope to uncover. Whereas traditional IT approaches follow similar processes as us, modern AIs don’t attempt to replicate how we think, but do it their own way. This approach is already paying off in many fields: for instance, AIs have been shown to be better at predicting heart attacks5 and some forms of cancer6 than doctors., Google is developing an AI program that reportedly detects 92% of tumours through high-level image scanning.7
The potential for AI in IoT is vast. IoT is generating vast quantities of data, much of which isn’t directly relevant for the purpose for which is was gathered. Think for example of an IoT application that gathers data on the condition of fruit while it’s in transit. It’s likely that it’s only the anomalies outside a range are actually of interest. The rest of the readings gathered are discarded as data exhaust. But that data, with its millions of tiny fluctuations, could hold all kinds of useful insights — perhaps even unlocking new revenue streams. The data might contain patterns that helps us prevent fruit spoilage, or predict the impact of warm weather in transit. We might not even think of looking for these correlations, but an AI might.
Getting the most from IoT
IoT is already delivering many benefits, but we’re only at the start. While individual IoT projects can be transformative, it’s only when IoT is managed as a companywide strategy that you really tap its huge potential. Many, many companies are now using IoT and making it a strategic priority. 29% of organisations globally and across all industries have adopted IoT, and 67% of adopters say IoT is mission-critical to their business.
Nearly half of adopters (46%) have already integrated their IoT projects with their core systems. A similar number (49%) are using analytics with their IoT data to support business decision making. If your organisation isn’t using big data and AI to get the most out of IoT data, you’re falling behind.
Over the next five years, we’re likely to see a massive increase in the use of AI and machine learning to analyse data, and turn it into actionable insights. There will be a shift from diagnostic data, towards more predictive and prescriptive analytics. As a pioneer in IoT, Vodafone can help your organisation stay at the forefront of this movement. Vodafone offers a wide range of end-to-end IoT solutions, with tailored analytics solutions to suit your business needs. We can help your organisation to capture and analyse the relevant data, no matter the size or scale of your IoT solutions.
Read the IoT Barometer 2017/18 for an in-depth look at how IoT is transforming the world of business—and what the future holds.
1 McKinsey Global Institute, The Internet of Things: Mapping Value Beyond the Hype, June 2015. ↩
2 Gartner, Gartner Says AI Technologies Will Be in Almost Every New Software Product by 2020, July 2017. ↩
3 Then called the M2M Adoption Barometer. ↩
4 The Washington Post, Zest Finance issues small, high-rate loans, uses big data to weed out deadbeats, October 2014. ↩
5 Engadget, AI can predict heart attacks more accurately than doctors, April 2017.↩
6 Cornell University Library, Detecting cancer metastases on gigapixel pathology images, March 2017.↩
7 Machine Learning, Google’s artificial intelligence can detect tumours with a 92% accuracy rate…, March 2017.↩