Share this post:

Searchable and actionable video content is king

Everywhere we hear that data is getting bigger and bigger and that one of the single biggest drivers of data growth is video.

Imagine – six years ago, 245 million (1) video cameras created 850+ PB of data daily. Today we generate videos in marketing campaigns, training videos, lectures, interviews, video production, Internet of Things (IoT) devices. Moreover, 62% of employees stream work-related video each day — such as training and development, company events, and town hall meetings — on smartphones, more than doubling those using desktop (27%). (2)

Considering above, I can only guess the number of data we are generating today having in mind the video quality has increased several times. IDC predicted that in 2020, there will be around 40 trillion gigabytes of data (40 zettabytes). (3) This immediately brings two questions: first, how do we manage such amount of data, and second, do we really benefit enough of this data?

Already established technologies like artificial intelligence and cloud computing, as well as the emerging 5G will and are already changing the video industry. For example, 5G will make mobile video streaming 10x faster not just for one individual, but for anyone streaming a video at the same time by creating a data super-highway.

But the question remains, what server resources do clients need to process and store all the video streams? How do they react to all the changes in their environments – new video devices, new compression options, higher resolution, regulatory requirements, etc. How to size efficiently the resources of a dynamic environment in scale? Should they find a new or expand their existing data center because of their video management systems?

This kind of questions I discussed with ECONT– one of the leading logistics services provider in Bulgaria that offers cargo, courier and postal services for e-commerce and individual customers. With more than 650 offices, ECONT has thousands of videos that they actively use to track parcels, protect assets, manage claims and optimize their work processes. ECONT was ambitious to improve even more the measures for handling data privacy of its customers and at the same time to be flexible in operating several petabytes of videos and make their video easy searchable, actionable and quantifiable. So, the company decided to make a step forward and to become an innovator in this industry by centralizing the whole video management system in IBM Public Cloud.

By having a centralized video management system in IBM Cloud, ECONT can enable rapid video review and search, real-time alerting and quantitative video insights. Leveraging data from video footage to quickly identify and respond to potential risks or opportunities and ensuring their quality standards.

For example, ECONT can quickly track the parcels throughout their way from the dropping point to the handing over in their facilities and resolve any disputes. The personal data and privacy of the customers is ensured with strict access control and anonymizing tools in the video management system provided as a service by IBM. In addition, Econt pays only the resources needed at the specific time, there is no risk of video loss due to infrastructure outages or burglary cases (the video recorders are the first thing burglars take) and there is no need to invest in and operate a huge IT environment on its own.

Video management in the cloud is becoming the new norm and the IBM Cloud is uniquely positioned to offer the highest security and bare metal servers with GPUs that can scale at any time. Video processing requires pure processing and cost-efficient storage performance – any virtualization just adds up cost and performance overhead – but what opportunities does it also create?

Most stored video goes un-analyzed because looking at all of it is humanly impossible. But even if we put a human behind every two cameras, human error increases drastically after only 22 minutes of monitoring two screens. And this is where AI can play a significant role.

Today video analytics can help us drastically increase the value of video data. Video Analytics empowers organizations to analyze and identify events and objects in real time and make this huge amount of unstructured data searchable.

We can now automatically spot security events at the moment when they arise and take immediate actions. But these capabilities have huge potential for the line of business (LoB) in your organization as well. Starting from uses cases such us KYC (know your customer) and customers’ behavior to custom models specific for your business needs.

IBM was one of the first companies to develop such a solution based on deep learning algorithms trained on millions of samples to turn video into actionable insights and transform the industries. But what really make a difference – IBM tools have the capability to train your own custom models with intuitive drag-and-drop interface without wasting the time of your AI experts and developers. This is a game changer that opens the door for gaining competitive advantage by using your creativity and business acumen. Your teams are getting actionable insights to act. And let’s not limit ourselves to video management, there is much more video content to explore in industries such as media, eSports, agriculture, transportation, energy and utilities, just to name a few.

………………………………………………………………………………………………………………………………………………………………………………….

1) IHS, Market insight,  https://bit.ly/2ZrjLGK

2) IBM Watson Media, Streaming Video Blog, Video Trends and Studies, https://ibm.co/3fxgcnR

3) Techjury, 25+ Impressive Big Data Statistics for 2020, https://bit.ly/2C3fikQ