Check out Optimize, our new AI-powered product!

Trends in Real-Time Communications
By: , ,
25 October 2016

How communication methods have evolved over time

Humans learnt the ability to speak and by extention communicate around 100,000 years ago according to the Science Museum. Over the centuries, the medium of communication has changed and adapted as the community size grew. For example, humans have used clay tablets, paper, oration, re-enactments, smoke signals, drums, printing, and more recently digitally over broadcast or the internet to communicate.

At the heart of the internet is our desire to communicate, webpages, tweets, blog posts, etc, all these are part of today’s digital conversation. This too is evolving and get more conversational: Bots, live streaming, video calls, and drones are part of the new trend. In this blog post, we will cover a few trends that we have observed over the past few months.

Live streaming

First live TV broadcast by BBC

First live regular TV broadcasts started in the early 1950s. News and sports events were the first programs to be broadcasted live over the TV networks. Throughout the period from the early live TV broadcasts to date, broadcasters have been looking for a way to spread their content as wide as possible and to add interactivity to the program. Finally, internet, computers and smartphones are offering a solution for both problems.

Live streaming over the internet is not a new thing, Ustream and Justin.tv began operations in 2007. Justin.tv transitioned into Twitch, which is a popular e-sport platform was acquired by Amazon in 2015. In the last year, several apps have appeared: Periscope by Twitter, Meerkat (recently shuttered operations), and Facebook Live. Most platforms doing live streaming today, let viewers interact with the streamer by commenting or send reactions while the stream is being broadcasted live.

In the 2016 season of the National Football League (NFL), Twitter has streamed NFL games live through its platform for free to all viewers. One issue that plagued the NFL stream on Twitter’s platform was delay, some users reported delays of up to 60 seconds, their neighbours were cheering before they knew what happened.

Large scale conferencing

Platforms capable of creating conversations between 100s of users. The typical use-cases involve live streaming to hundreds to millions of participants and then dynamically patching in the viewers to get their opinions or voice heard (airtime). There are several use-cases, for example patching in a student to ask a question on a online classroom or a shareholder during a Annual General Meeting.

Hangouts reportedly increased the max number of simultaneous participants in a conference to 25. Using other platforms, we observe some customers doing conferences regularly with 30-50 participants on callstats.io, the largest so far had over 110 participants. Large scale conferencing requires solid servers that can distribute the video signal to all participants.

We are expecting to see a growing number of products incorporate live streaming broadcasts, as there are more SDKs emerging in this space. For example, PhenixP2P, Jitsi, Kurento, and Red5 Pro. These platforms endeavour to reduce the end-to-end latency from a few seconds to 100s of milliseconds.

Video filters and effects

How Snapchat filters work

Introduced by Snapchat (now called Snap). People love these filters, they incorporate them in their storytelling or just whimsically. Many products seem to incorporate them: MSQRD, Instagram, and a host of companies selling filters as a service.

Flying cameras and surveillance with real-time video

Close-up filming of fireworks with a drone

Using drones for aerial video footage is an inexpensive method compared to a helicopter. Furthermore, since a drone is unmanned, smaller, more versatile, it can be flown in any environment with low risk of damages. The video above is a good example, a drone is used to film close-up of fireworks in the sky. There are not many helicopter pilots that would agree to fly in there.

Hexo+ and DJI, both have created self-flying cameras, essentially a drone that follows an object, typically, filming a person. For example, the Hexo+ drone can be commanded to hover close to an object, or to circle around the object, etc. Meanwhile, DJI’s drone can automatically fly around obstacles whilst following the person. Not so long ago, filming a sports action video costed thousands of dollars and nowadays it can be created without a filming crew and at a fraction of the cost. The kicker, these drones can stream live to an audience!

Entertainment aside, drones can be also equipped with different type of sensors. For example, SharperShape’s drone can inspect assets, such as forests, real estate, or power lines. The video (LiDAR, UV, IR) and the sensor streams the data live to the operator, allowing the inspections to be carried out faster and cheaper.

Ring.com is a doorbell with a camera, connects your front door to your smartphone. So that you can see who is at your front door in real-time, even when you are not at home, say on a vacation 1000s of miles away. In case you are unavailable, it records the video activity for perusal later.

Bots

Earier this year, Microsoft announced a bot framework in March, and Facebook revealed theirs too. In August Facebook announced that are more than 11,000 bots on their Messenger platform and they have a bot lets you talk with the president of the United States.

Furthermore, there are a growing number of bot frameworks in the market: Wit.ai is a popular bot engine that has over 23,000 accounts. Pullstring, the company behind the talking barbie, offers a platform that hooks up NLP and speech recognition technology for developers to build conversational bots. Slack announced their bot-kit for their platform.

From the WebRTC perspective, a bot worth mentioning is related to real-time communications is Clarke.ai, it joins a conference calling in like a real person and makes notes automatically of the call. Zendesk launched a new product Zendesk Message, which combines messages from bots and support agents. The conversation is first initiated with a bot sending information to the customer, and an agent takes over when the customer requires more individual care.

There are other signs of the bot industry gaining more momentum. At Disrupt San Francisco 2016, we came a across a couple of bot analytics companies: dashbot and botanalytics, they both provide analytics around the performance of the bot (such as, engagement).

In summary

The future of real-time communcations looks very bright. In addition to the products and services mentioned in this article, devices and software for augmented and virtual reality are going to become big trends. Once matured they will also add new dimensions to real-time communications. If you are as excited about these real-time communication products as we are, come work for us.


As a disclaimer, the companies and products are mentioned in this article because we think they are doing something interesting.



comments powered by Disqus