Video analytics can feel like a taboo "Big Brother" subject, but it's a technology that already exists and is only going to play a larger role in our lives going forward. We talked to Blueprint Managing Director of Innovation and Video Analytics expert Gary Nakanelua to break the stigma and discuss how your business can leverage video analysis efficiently and responsibly.
What do we mean at Blueprint when we talk about video analytics?
Video analytics is really about looking at video as a data source and converting that into something that can be integrated into analytics projects.
In the retail environment, for example, most companies already employ some type of video capture – usually for lost prevention purposes. They use it to help the authorities if an incident happens and an investigation occurs. What that results in is a very visual interaction with video. If you think about it a little differently – you can convert that video into a malleable data source that you can do other things with. Yes, you can identify a particular individual, but you can also capture every person walking into the store. You can find out how many people are entering the store, what path they are taking through the store, how long they are spending with employees, do they pick up and interact with items, are they wearing a mask. Retail stores are already capturing all these behaviors on video, which is raw data. But they need to go further and convert that footage into something that can feed into analytical processes.
What makes Blueprint's perspective on video analytics different?
Video analytics is not new in and of itself, it’s been around a long time. Most video analytics companies require a significant investment in hardware to even begin, but Blueprint doesn’t, we meet customers where they are. A retailer, for example, already has camera infrastructure and we work with that infrastructure. Depending on the specific situation, some adjustments may need to be made and we would have conversations about what can be done with what the company has in place, but it generally doesn’t require they make a significant investment in hardware. Video analytics companies that approach a customer with an off-the-shelf solution “that can solve all your problems” ― it just isn’t true. We’ve learned that video analytics is very much a solution-based approach that needs to be tailored for the environment and for the customer. You need to be able to adapt to their needs and what they have in place already.
What are the major components of an effective video analytics project?
As much as video analytics is an established discipline, it isn’t exactly commonplace, so customers often don’t know where to begin. It starts as a conversation to help identify what they don’t know ― exploring the art of the possible. We will also have to have a technical exploration. We go through what the organization has, maybe that is surveillance cameras, and we explore what can be done with that data. It’s like dumping a big bag of Legos on the ground and picking through it to see what you can make.
Then you look at the specific industry and what information the organization needs. In retail, for example, the count of people who come in and out is very important because customer count can be used to determine the average sales per customer and to understand customer traffic to better optimize employee scheduling.
So, a customer may know why they originally set up a camera infrastructure, but they don’t know what they can do with it. By having these conversations, you start to uncover what is important to them and how you can create a unique solution for them, often with the equipment they already have in place.
What are the benefits of getting video analytics right?
There are several things you can do with video analytics, you can identify people, plot out paths people or vehicles take, you can create heat maps of things. There are many things we can do, but the question really is, “what does the customer want to do?” It’s about finding the data that well help them optimize their business and have that quick return on investment. However, it’s a journey. Video analytics will not provide an instant increase in sales nor will operational deficiencies immediately decrease. Those things will happen, but you must dial it in. You want to develop the infrastructure in a way that allows you to adapt and customize it to your needs as you discover them and that adaptation needs to come at a low cost.
What are the risks companies run by not investing in video analytics?
The main risk is that an organization would not be taking full advantage of an investment, in this case video-capture hardware, that they’ve already made. They have raw data that they are capturing, but in effect they’re just hoarding it. At some point the decision to add video capture capabilities was made, so now you have all this video equipment and footage, but nothing to show for the investment. That’s not even mentioning the ongoing maintenance costs. These sorts of investments tend to look good on paper, but it turns out they do not impact the business in a positive way unless the video captured is fully utilized to drive the business. And because of that, you miss out on the possibilities at your fingertips and risk being overtaken by the competition.