When you clicked on this article, you triggered a query to see the page. If you had a good connection, it loaded almost immediately. It was fast, but was it real time?
Most people would answer yes, but given the definition of real time, the answer is no. Regardless of how fast the article loaded, it was written long before you queried to read it. Since the source was not generated immediately, it’s not considered real time.
Conceptually, real time refers to the production and transfer of information at the speed of experience–that is, as fast as the present moment.
In other words, the information, image, video, or sound is produced and delivered as fast as it would if we were talking face to face.
That said, the term “real time” has taken on different meanings as it’s been applied to different contexts. This article defines real time and discusses it in the contexts of:
- Live entertainment
- Operating systems
- General systems
- Marketing, and
- Software updates
I’ve also done analysis to see “how real-time real time actually is” by examining the speed of delivery in contexts such as Instagram and Facebook likes, Robinhood order placements, Google Earth updates, and other popular use cases.
Out of Scope: Real Time with Bill Maher & Polymerase Chain Reaction
If you’re looking for information on the show Real Time with Bill Maher, it’s not the scope of this article. Instead, check out the HBO network site for more information.
This article also has nothing to do with real-time polymerase chain reaction, which refers to a data collection technique in mRNA quantitation. While there are similarities, this article focuses on information technology and business, not natural sciences.
Real Time Definition
Real time refers to the delivery of information through a system as close to the speed of experience as possible–that is, information produced, transferred and received as fast as a face-to-face conversation.
For example, Instagram “likes” are considered real-time because a “like” appears almost instantly. Likewise, real-time computing refers to systems that respond within milli- or microseconds. The delay is insignificant, and it’s nearly as fast as a conversation, which makes it real-time.
“Real Time” vs “Real-Time”
Cambridge Dictionary, Dictionary.com, and Collins Dictionary all recognize both spellings. The difference is that “real time” written without a hyphen is a noun, usually preceded by the preposition “in.” Written with a hyphen, “real-time” is an adjective.
For example, you would write “this is a real-time computer” and “this computer processes information in real time.”
Real-time computing is a general term for software programs that must provide output within a user-defined period of time following input. In most cases, these response times are in microseconds. For example, online trading platforms must register transactions in microseconds to guarantee a price.
Real-time computing is dependent on several other real-time systems, including real-time operating systems, synchronous programming languages, and real-time networks. Each of these provide the correct framework on which real-time computing can be executed.
Real-time operating system
A operating system is the lowest-level of software on a computer that controls the unit’s hardware. A real-time operating system is one with real-time computing constraints that are compatible with the development of real-time software.
Synchronous programming languages
Synchronous programming languages are those that interact constantly with their environment, rather than receive an input, process, and communicate an output. The earliest synchronous languages, Esterel, Lustre and Signal, were invented in France in the 80s. We won’t go into the details of how they work — just know that they’re essential to real-time computing.
While operating systems and synchronous programming languages are found within a single computer, networks are by definition a collection of computers. Real-time networks are nothing more than webs of connected computers that all have real-time OSs and languages.
Distributed and Embedded Systems
At a higher level, real-time networks can be viewed in terms of two systems: distributed and embedded systems. Distributed systems are groups of computers all aimed at accomplishing a common goal. In other words, they’re comparable to networks.
Embedded systems, however, are independent hardware units within a computer that perform a function aimed at a goal common the computer and other embedded systems.
In the context of real-time computing, the common goal of distributed and embedded systems alike is real-time processing.
Other than computing, so far we’ve discussed real time primarily in the context of entertainment. Real-time data is similar to streams and live feeds, but rather than consumer content, the transmission is data.
Real-time data is information that’s delivered immediately after it is collected. A great example of this is ad placements online. Ads are placed based on an optimization algorithm that considers prices, targets, and time constraints.
A good example of data that’s not real time is Google Earth data. While it feels like Google Earth is always up to date, the satellite images are processed many time over before uploaded to the site.
Just like real-time computing requires operating systems, synchronous languages, and networks, real-time data has technical requirements. Active databases allow real-time data to be communicated continually from input to output.
Real Time Vs Batch
A similar but distinctly different type of data processing is called batch. Batch processing collects groups of data in a user-defined size, then processes these groups and delivers the output. We mentioned before that Google Earth is not real time. That’s because the data is processed in batches — after the fact.
How Real-Time is Real Time Actually?
Before I address other important comparisons and popular contexts, it’s time to address the elephant in the room. Real time is supposed to be immediate, but in reality nothing is as fast as the speed of experience. Which begs the question: how real-time is real time, really?
To answer this question I performed a few experiments and collected data on 5 common examples of data transfers that we often imagine happen in real time–that is, where information is produced and received almost immediately. The five examples are Robinhood stock orders, Google Business Information changes, Likes on Instagram, Likes on Facebook, Google Earth updates.
Robinhood order placements during normal trading hours (9:30 a.m. – 4:00 p.m. ET) are real-time in most cases. I did an experiment with 10 orders on low volume shares and found 100% of them were booked instantaneously. This means the Robinhood technology can indeed be considered real-time.
That said, Robinhood’s stock order routing requires there be a buyer and seller on either side of a trade. This means that a high-volume trade on a low-order stock could result in a delayed transaction time.
Google Business Information
When you’re looking for businesses on Google Maps, it may seem like most businesses’ information is up to date. However, in some places businesses aren’t active, and you may not see the listing at all. These owners need to update their information.
Updating isn’t always immediate, and in some cases Google does not allow you to make changes since Google sources the information elsewhere (such as reviews). As a whole, this means google businesses and maps are not a real-time software.
Likes on Instagram
To test how real-time likes are on Instagram, I Liked four of my friend’s pictures and times how long it took for the notification to appear on the picture. Believe it or not, the average delay is 2.75 seconds. This delay is considered real time, but I for one thought it was faster.
Likes on Facebook
Likes on Facebook had a similar delay to those of Instagram. After testing four likes, the average delay was 2.29 seconds. Just like Instagram, this is considered real time, but the speed will certainly improve in the future.
Google Earth shows satellite and street-by-street images of Earth. The process of collecting this data requires that Google drive a car on literally every main road on Earth. In addition, Google needs to process the images thereafter in order to blur out people’s faces and avoid showing sensitive personal information. Because of this delay, Google Earth is not a real-time system.
Real Time vs Live
Live refers to the the upload of information for diffusion as soon as it occurs. The most common examples are live radio and TV programs. However, when these programs are not delivered immediately, they’re not real-time.
If you’ve ever watched a live sporting event in a bar, you probably regretted it when people watching a different TV yell before a play ends. There are varying delays in the broadcast because it’s not instantaneous. The upload time is live, but the delivery of information is not in real time.
Some live feeds are indeed real-time. They’re improving continuously, and the fastest provider will sweep the market in the future.
Real Time vs Streaming
Streaming refers to a continuous flow of data from a source. While we most often hear it used in the context of video on demand (like Netflix and Disney+), streaming also applies to live entertainment. As long as the data flow is continual, it’s streaming.
However, video on demand is not real-time because it is not produced and delivered at the same time. Just like this article was written before you clicked to read it, so too is video on demand.
In other words, if streaming delivers content immediately produced, such as live events with no delay, it’s real time. Otherwise, it’s not.
Just like streaming, since TV is pre-recorded, it’s not real-time. Note, however, that in film production, the cinematographic attempt to portray events as they occur is called the “real time technique.” The show 24 attempted to do this. But that’s a little out of scope of the article…
Real-Time vs Near Real-Time
“Near Real-Time” is a term used to address the reality that real time is never actually real-time. As seen with the examples of Instagram and Facebook Likes, the fastest processing available today may be microseconds, but it is never 100% at the speed of experience.
You may hear some one say a computer process is “near real-time.” In most cases they’re referring to the delay due to human interpretation of output data of a real-time system.
For example, live survey results in a live studio may be collected in real-time, but because an announcer needs to interpret them, the whole process is “near real-time.”
An interesting innovation was real-time text. In messages, real-time text is the transmission of each character as it is typed, regardless of whether the sender authorizes the message. Today, it’s primary use is on-screen captions written from the spoken word for those who are hard of hearing or deaf.
Real-time payments are likely to upset the payments industry in 2023, when the US Federal Reserve’s real-time solution will become available. But what are they?
Real-time payments are money transfers that happen almost instantaneously, rather than taking a few hours, or even a day, to move money from one account to another.
While social payment apps such as Venmo are instantaneous within the app, it’s not yet possible to transfer the money to a bank account instantly.
And if they money is not in a bank account, then it’s not useful for daily purchases. In addition, social payment apps usually have a transfer limit, such as $300.
So real-time payments available in bank accounts will surely be a revolutionary innovation. Hats off the US Fed.
A dashboard is a single page view of multiple metrics. Originally, the term “dashboard” was used in reference to the dashboard of a car or plane, where metrics such as speed and engine temperature were available. And this was real-time.
Is the digital world, dashboards evolved to show the performance of subjects such as websites or manufacturing performance.
But in many cases, these dashboards were not real-time. Instead, analysts drew down data and built visualizations that they copy/pasted into a dashboard view.
A real-time dashboard is one in which the data source for the metrics is directly connected to the collecting source, which is made of real-time data. A common tool for real-time dashboards is Tableau.
Real-time marketing uses real-time data to create bespoke content and solutions for customers. For example, if there’s unexpected rain in an area, apparel marketers can quickly adjust to promote rain jackets on social media. And when it stops raining, to promote boots.
At a very high-level, the concept of a “real-time update” refers to any receipt of information occurring immediately that allows the receiver to take relevant action. You can apply this to almost all of the contexts discussed above.
If you found this article helpful, feel free to check out more free content on data, finance, and business analysis at the AnalystAnswers.com homepage!