What’s next for IoT?
There’s no doubt that there’s a lot more “connected things” these days and that means a lot more data. Specifically, technology is moving out of the consumer’s hands and into Healthcare, Oil & Gas, Transportation, Aviation and more. The spread of smart devices and sensors creates new forms of value and brings challenges for enterprises seeking to exploit this technology. However, while this boom in data has the potential to advance the industrial space in ways never before thought possible, few companies today have the right technologies in place or the right business models that can truly utilize the power of IoT.
So what should enterprises know about the permeation of sensors and connected devices and how can they prepare now or risk being left behind by their faster moving competitors? Here are 4 things to keep an eye on:
Smart Meters for a Smarter Planet
Consider smart meters’ (the kind that Talend Data Masters’ winners m2ocity and Springg are using) impact on data growth. Each single device typically generates about 400MB of data a year. Not much, right? Well, the numbers add up quickly! According to a recent Bloomberg report, it’s predicted that 680 million smart meters will be installed around the world by 2017 leading to an estimated 280 petabytes of data generated by smart meters each year!
How can this lead to a smarter planet you ask? Consider Springg, an international software company specializing in agricultural data transport, software development and sensor technologies. Utilizing Talend software, Springg is able to evaluate data collected from sensors used in its mobile laboratories around the globe to measure soil elements. Based on these insights, Springg can give farmers soil and fertilizer advice to help dramatically increase their yield targets—thereby helping those farmers feed the world.
Data Up in the Sky
Watching for another area of machine-to-machine data growth? Try looking up in the air. A report from Wikibon found that the aviation industry is ripe for innovation in the form of Big Data analytics. An airplane flight (depending on length) can generate up to 40TB of data that will be used to analyze flying patterns, tune jet engine functions, identify new routes and reduce downtime. This is the type of data-driven decision making that has been revolutionizing eCommerce. Airlines with the ability to handle their big data in real-time and turn that data into instant will ultimately gain the competitive advantage.
For example, Air France/KLM. Each of the airline’s A380 aircrafts contain roughly 24,000 sensors, which generate 1.6 GB of data per flight from Smart meter technology. The company utilizes this data to detect breakdowns before they occur. Using the smart meter analytics technology, Air France can now detect potential needed repairs 10 to 20 days before they occur. This prevents immobilization of the aircraft, which is not only expensive for the company, but also impacts their overall level of customer service and revenue.
Machine Learning and Data Science
Data is the life-blood of your business. You need to not only absorb as much as you can, but you need to analyze it and find any secrets. Data science and machine learning are gaining adoption as companies become even more data-driven. Machine learning is a form of artificial intelligence, where computers can learn and act to make decisions – sort of automating some of the data science tasks. With the shear volume of data coming from the billions of Internet of Things (IoT) devices, automation is a good thing! Spark MLlib is gaining popularity with its many machine learning algorithms for customer segmentation, forecasting, classification, regression analysis and more. Incorporating machine learning into heavy data loads will be an important step for companies to make better use out of the wave of information coming from connected devices – and find the needles in the IoT haystack.
Even though we now live in ‘the Golden age’ of big data, what will likely surprise you is that most companies are NOT utilizing the true value of the data to its full potential. In fact, a recent study by McKinsey showed that less than 1% of all IoT data is actually being used for decision making today. WHHHAAATT??
Why you ask? Mostly since the IoT data is being used for alarms or real-time control, not really optimization, or predictive and prescriptive analytics. Also, there are many challenges in making machine learning a reality. Data has to first be organized and cleansed. It can take a long time (months!) to put a model into production—particularly when analytics models change frequently (requiring more updates), and there is a lot of hand-coding going on….Not exactly in the best interest of the data-driven firm. But there is a solution! Companies applying open, native technologies like Talend for real-time big data integration, which requires zero hand- coding, utilizes Spark, Spark Streaming and Spark machine learning, can start to get more insight from their data.
Want to learn more about new capabilities in Talend 6.1 for machine learning and real-time big data? Check out our free on-demand webinar on December 17th and automatically be entered to win two movie tickets to ‘STAR WARS: The Force Awakens’, due out this week! Register here.