Check out the on-demand sessions from the Low-Code/No-Code Summit to learn how to successfully innovate and achieve efficiency by upskilling and scaling citizen developers. Watch now.
That’s the number of IoT-connected devices expected by 2025, and each of those devices will create its own stream of data. In this new landscape, it’s not hard to see why time series data is becoming one of the most valuable commodities in technology.
At the most basic level, time series data is any data that is organized and sequenced by timestamps. The sources of this data vary; in the physical world, devices such as temperature gauges, sensors, or batteries generate it, and in the virtual world, it comes from software, systems, microservices, or virtual machines.
As we witness the “sensor-fication” of the world — from machinery on the factory floor, to self-driving cars, to solar panels on your roof — the data those devices produce becomes the key ingredient in digital transformations. To effectively harness and analyze this data, organizations utilize time series databases, which have the ability to handle the unique workloads required of time series data. For instance, a time series database can handle high-speed data reads and writes, and at higher volumes than other database types.
Intelligent Security Summit
Learn the critical role of AI & ML in cybersecurity and industry specific case studies on December 8. Register for your free pass today.
Here we will take a closer look at what time series data is, what it can do, and how businesses and organizations of all types and sizes use it to change the world as we know it.
Time series data: A primer
A sequence of data points indexed in order of time is the essence of time series data. A series, therefore, consists of successive readings or measurements from the same source over a time interval to track changes over time.
There are two types of time series data: “Metrics,” or measurements gathered at regular intervals of time, or “events,” measurements gathered at more irregular time intervals. Time series metrics are great for forecasting because historical data is a solid indicator of what could happen in the future. On the other hand, we find time series events in things such as event logs, which can provide useful information on network traffic, usage, and much more.
Time series data is ordered, making it unique in the data space because it often displays “serial dependence,” which occurs when the value of a datapoint is statistically dependent on another datapoint from another time.
What it can do
Time series data provides the context of time. This context is both critical and valuable because everything happens at a particular point in time; if you know when something happens, you can make informed decisions about them, such as when to act or when not to. The data can also be used in time series analysis for a wide variety of cases, including (but not limited to) tracking stock prices, forecasting sales figures or healthcare monitoring.
There are two main reasons behind the surge in this type of data, the first being the rapid rise of connected IoT devices. A single sensor pumps out loads of time series data. Now, imagine a device with 50 sensors. Further, imagine a company selling millions of those devices. With nearly 30 billion connected devices expected in just a few years’ time, it’s easy to see how quickly the volume and scale of time series data increases in the IoT space.
The other reason is the way organizations use the data. As businesses move significant portions of their data to the cloud, the systems, containers and processes involved all create time series data. This, in turn, allows organizations to reuse their data across continually expanding networks.
What time series data does at a practical level
At our current trajectory, everything that can be instrumented will be instrumented, making it easier than ever to obtain information on the state of the physical world. As it stands, we already have the ability to leverage these streams of data with AI and machine learning (ML) to create insights that allow us to act quickly. The current state of self-driving cars, traffic navigation, and “smart” buildings and appliances provide a glimpse into what the not-so-distant future will look like.
The creation of data certainly isn’t limited to the physical world. It seems that, with the rise of microservices, containers, and serverless IT architectures, instrumentation in the virtual world is growing just as fast — if not faster — than it is in the physical world.
The use cases for this data will no doubt continue to grow, but plenty of companies already use it to make the world a better and greener place. Three examples include:
- Sustainable energy utility company Bboxx uses a time series database to build its proprietary operating system, which monitors thousands of solar panels and batteries across the developing world, specifically Africa. Now, more than 2 million people have access to clean energy from homes powered by Bboxx, thanks in large part to the company’s innovative use of time series data.
- In Belgium, wind energy pioneer VLEEMO (an abbreviation for “Vlaamse Ecologie Energie Milieu Onderneming) uses time series data to help monitor shadow flicker caused by wind turbines as their blades rotate. These types of moving shadows are known to cause stress and headaches, and the country limits the amount of shadow flicker allowed to 30 minutes per day. VLEEMO also uses this data to monitor the accumulation of ice on turbine blades, which can be potentially dangerous to nearby people and buildings. Collecting all of this data helps the company maximize the energy output of its turbines.
- Finnish company EnerKey ingests and combines raw data to analyze and forecast energy usage in real-time. In a basic example, the company correlates weather data (which is also time series data) with energy usage data to predict the energy needs of individual buildings. This process can help energy companies save hundreds of dollars per month per building, resulting in huge savings when considering the hundreds or even thousands of buildings the company serves.
Ubiquitous, but still growing
As time series data becomes more prevalent, expect the process of manually monitoring physical dashboards to disappear. AI and ML can (and in some cases, already do) monitor data-driven trends and react automatically based on pre-defined rules. Ultimately, this frees up teams from relatively mundane work and allows them to innovate even further.
Time series data is already ubiquitous, as it lies in every part of today’s digital businesses, but it has yet to reach its full potential within most organizations. Yet as the amount of data generated continues to grow, those who become adept at harnessing it, analyzing it, and using it to make critical decisions give themselves the best chance to create competitive advantages.
Evan Kaplan is CEO of InfluxData
Welcome to the VentureBeat community!
DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.
If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.
You might even consider contributing an article of your own!