« Updating Bank Technology Results In Rediscovery | Main | To Create A Global Enterprise, Start With Employees »

August 19, 2015

Why Algorithms Are In Everyone's Future

Posted by Dr. Srinivas Padmanabhuni (View Profile | View All Posts) at 9:40 AM

Stanford engineering students teach autonomous cars to avoid obstacles [Source: https://www.youtube.com/watch?v=NNIFNmpB6Oo]

Recently, I was reading a new book about the Internet of Things (IoT) and I was struck by something that the author had written: As markets grow larger and more international, the segmentation of those markets will become all the more precise until each market is segmented to one consumer.

One person! We can thank algorithms for this kind of ultra-focused segmentation. To the average person on the street, an algorithm is nothing more than a mathematical recipe of sorts. But large enterprises are learning to write algorithms that can account for human biases, thereby making the machines that run them more lifelike than ever. We live in an age in which algorithms rule. No longer will senior executives run companies on gut instinct. In the near future, algorithms will be at the core of deep analytics for just about everything: The IoT, financial trading, and, yes, even the driverless car.

Remember how some people had an innate sense of where the capital markets were headed? They would invest their money accordingly and often make big windfalls if stocks rose or crashed. But increasingly those tasks are going to algorithmic trading models that take advantage of the slightest swing in the price of a stock. No gut feelings required.

I was fascinated by the story about Google employees who were working on the very ambitious Google Maps project some years ago. Sometimes the photographs were a little grainy and the employees had a difficult time figuring out whether the images were street numbers on houses or just graffiti. Some engineers at Google wrote some algorithms that turned into what's now known as the Google Brain. It reads the numbers off the photographs of houses and (as you might imagine) has many other uses.

Think about how algorithms are changing the way we travel. Today, the car's computers can anticipate what is transpiring on the road ahead of it faster than a human driver can. Many of us used to think of connected cars and houses as part of a neat home contraption. You could potentially control the temperature in every room of your house as you sat in your office many kilometers away. Or you could unlock the doors to the house or car using your smartphone. The truth is, the Internet of Things has become a lot bigger than just an enabler of home appliances because of the sophistication of the algorithms that run it.

Machine learning algorithms are now a reality. They have already taken control of some industrial systems and have had wonderfully efficient effects on global corporations. Instead of a machine speaking to a manager, a machine talks to another machine - one that is automated and is reasonably intelligent.
 A colleague of mine has said that production lines can now operate for as long as possible and consume as little energy as possible, thanks to machine learning algorithms. They seldom break down and produce items of high quality with the utmost precision. 

Sensors and control systems on production lines that utilize machine learning algorithms can capture vast quantities of data. When these systems are scrutinized with the proper analytical software, it is possible to assess how efficiently and precisely a high-speed cutter mills components. Then, with the benefit of data analysis, manufacturers can optimize their systems to produce less waste, consume less energy, and improve the quality of their products.

Now that all companies, no matter what industries they might be in, are making decisions based on algorithms, their data must be real-time to have the intended effect. Failure to use such tools will cost enterprises dearly. The ones that do use these new 'super-algorithms' and methods effectively will be the market leaders of tomorrow.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.

Search InfyTalk

+1 and Like InfyTalk

Subscribe to InfyTalk feed

InfyTalk VBlogs: Watch Now

Infosys on Twitter