Who Was Smarter, Karl Benz or Sigmund Freud?

How’s your European history? Today, I’m going to take a brief look into the worlds of two of the leading minds of the 19^th century. Two people that fundamentally and indisputably influenced how we live today. We’ll consider how different their domains really are and what that can teach us about the new field we could call analytics of things: delivering valuable new business outcomes with IoT and other asset-related data. A stretch you say? Not at all…
 
Let’s start with Sigmund Freud. He was born in Austria in 1856 and qualified as a Doctor of Medicine. He’s the founder of psychoanalysis. He and his theories are often misunderstood, but there’s no arguing against the fact that Sigmund changed forever our understanding of how we think and behave, and the drivers behind those things. Love him or hate him; subscribe to his theories or not; his work “still represents the most coherent and intellectually satisfying view of the mind” according to Nobel Prize recipient, Eric Kandel.   
 
Karl Benz was born in 1844 in what is now Germany. He was an engineer, designer and inventor. Most relevant to this discussion perhaps, he is considered the inventor of the first practical automobile (when Henry Ford was around 2 years old).
 
Both men were geniuses, and both played a massive part in defining our modern world. However, what’s that got to do with analytics? Or outcomes? Or the IoT? Put simply, most of today’s analytical techniques and the answers they deliver fall squarely into only one of these men’s domains. However, as we move to the era of the Internet of Things, that is going to change. This is because Sigmund’s domain of people and behaviours and Karl’s domain of assets and the laws of physics are very different indeed.
 
We use analytics to study and influence people every day. This is Sigmund Freud’s world. Businesses and governments use statistical techniques to understand how people think, predict their behaviours and suggest how best to respond. We use analytics to nudge a group or individual’s behaviour towards buying more products, or to follow a safe and efficient route home from a crowded event. We also use analytics to predict fraudulent behavior; or to understand which subscribers, customers or citizens are key influencers. When we understand who the key influencers are, we can reward their loyalty, encouraging them to promote brands and services for us.
 
In the industrial IoT world - the modern evolution of Karl’s world - it’s very different. In this asset domain, statistical probabilities can be valuable. But in many cases, the fundamentals we understand about analytics in retail, banking, insurance or in any B2C, people-related industry simply do not apply. The statistical models - even the skills and capabilities - we utilize to understand and influence people aren’t always going to do so well, addressing “things”.
 
We can’t persuade a pump to perform better. We won't ever find a  subset of electric motors is choosing to lie to us. There is no way to improve our relationship with our transformer fleet to make sure the most valuable ones never churn from us to another utility.

 
Another complication for traditional analytical techniques is that IoT data sets tend to have a finite and well-understood set of possible values. Statistical outliers probably more often signify a faulty sensor sending nonsensical readings than any new critical insight. After all, as Montgomery Scott, chief engineer of the Starship Enterprise, was known to say, “ye cannae change the laws o’ physics”.
 
As such, the analytics of things is often about complex mathematical models, not just stats. It’s about interpreting detailed time series data, recording constantly varying voltages and frequencies and flow rates. It’s about correlating that operational information with traditional enterprise data from supply chain and human resource systems. Analytics of things is even about helping businesses cross the great cultural divide between those that own and manage operational technologies that keep plants running and the IT teams.
 
The analytics of things is about working with - and being credible with - people that have an entirely different mindset to the marketing director or the head of sales; people like maintenance directors and chief engineers that often have no real concept of “the end customer”.
 
Sounds hard, right? And if it’s so hard, why wouldn’t we just stick with the people and behaviors domain we know so well? Simple: there will be more than 55 billion IoT devices by 2025, up from about 9 billion last year. There will be nearly $15 trillion in aggregate IoT investment between 2017 and 2025.  The Industrial IoT isn’t about the things, it’s about the new insights that will transform your business. Analytics of things is the future of analytics. It’s really as simple as that.
 
Way back at the start, I asked who was smarter:  is it Karl, in his asset domain or Sigmund, in his people domain? There is no definitive answer. The much more important point is to understand that these two geniuses operated in vastly different domains. When it comes to analytics, to data science, to delivering data-driven business outcomes, it’s important to understand that the analytical skills and techniques that we have honed for so many years, often don’t apply to the analytics of things.
 
These skills don’t make Karl smarter or those of us at Teradata with engineering degrees smarter than those with psychology degrees. They make us different. We learned different things at school.  If we - and you - are to do as much with enterprise-wide analytics in the future Industrial IoT world as we have in the world of people, we need only recognise, embrace and make the most of those differences.

 We’ve been doing the analytics of Sigmund for years. Analytics of things is the next great data revolution. We need to be ready for the analytics of Karl.