The number of connected devices and assets that comprise the Internet of Things is nearly unfathomable. Last year, Gartner predicted there were more than 8.4 billion connected “things” – roughly a billion more than the entire human population. While these numbers vary by analyst and media outlet, there’s no denying that this growth will continue as companies in nearly every industry look to cash in on this mega trend.
However, the number of connected things aside, the IoT presents a huge business opportunity across a variety of sectors. In fact, according to IDC, spending on the IoT will reach $1.2 trillion by 2022, predominantly in technology and services. For example, equipment manufacturers will invest to improve the customer experience and minimise service disruption, while governments will invest to intelligently monitor and improve the ageing grid and infrastructure. Essentially, all industries will capitalise on the IoT by creating entirely new business models based on the insights gleaned from its data. Simultaneously, well-funded, data-first start-ups will to go to market even faster than traditional companies in those industries with disruptive products and services.
Managing unprecedented data growth
Billions of connected things generate unprecedented volumes of data. According to recent research, 2.5 quintillion bytes of data is created each day. Even more astoundingly, 90 per cent of the world’s data has been generated in the last two years alone.
Initially, this data growth presents more challenges than benefits for traditional enterprises, particularly when it comes to managing and governing this information. Why? Because they need to re-platform and modernise their data warehouses to take advantage of modern approaches to streaming, managing, and analysing. This applies not only to sensor data, but all forms of data.
On the other hand, data-first companies born in the 21st century can make a first-mover advantage with a green-field approach. By quickly evaluating and adopting proven technology, they are able to form their analytical stack without the baggage of data infrastructure.
Both established and emerging businesses have been building up data lakes for years – storing sensor and other emerging data types in cost-effective ways such as Hadoop and S3. These organisations were waiting for a time when IoT business cases matured far enough for them to apply analytics to their data at scale, enabling monetisation. And that time appears to be now – as technology has advanced, enterprises and start-ups can now focus on capitalising on stored data, creating new revenue streams as a result.
Why successful IoT use cases require predictive analytics
Predictive maintenance. Smart metering. Pay-as-you drive insurance. Telemedicine. Intelligent manufacturing. The list of emerging and proven IoT use cases is growing daily. However, any organisation implementing one or more of these use cases as a strategic part of their business requires one essential technology – analytics, or, more accurately, predictive analytics. Hiring legions of data scientists to build machine learning models in Python or R is not enough. These IoT use cases require highly accurate and constantly updated predictive models based on unlimited volumes of data – not just samples of the data.
To provide an example, Nimble Storage sought to differentiate itself in the storage industry through achieving higher levels of customer satisfaction than its more established competitors. To accomplish this goal, the company invested in a high-performance, massively scalable analytical platform that manages and analyses data from millions of sensors to predict and prevent any potential indicators of downtime.
The initiative was so successful that Nimble Storage essentially made storage autonomous, predicting and preventing 86% of issues automatically. The company was ultimately purchased by Hewlett Packard Enterprise for more than $600 million, as it hopes to apply this proven predictive maintenance method across its full portfolio of storage and server offerings.
Business drivers require action for analytical insight
There are three primary business drivers behind IoT use cases. Two of these focus on improving customer satisfaction through less downtime or disruption, while at the same time reducing the costs often related to service, support and maintenance by automating a business process. However, the most impactful business driver revolves around generating entirely new sources of revenue based on the inherent value of the new connected product or service.
To pursue these business drivers, equipment manufacturers, governments, industry leaders, and just about any other organisation in the world will invest in IoT analytical technology this year and beyond. They will need to consider new streaming technologies, more scalable, open analytical platforms with machine learning built in, and potentially edge-based analytical offerings.
Whether it’s a modernisation initiative or a new IoT project with a fresh approach, organisations will only be successful if they can access predictive insight in mere seconds or minutes from unlimited data volumes, allowing them to uncover trends and patterns before the competition. To be truly disruptive, IT leaders will need to change their mind-set and look beyond cost reduction and operational efficiency. This way they can create entirely new businesses where revenue streams are driven by data analytics.