Transform what video intelligence can become with AI

AI-powered video analytics are turning security cameras into robust sensors that can improve how you monitor security at your sites. With artificial intelligence (AI), cameras can detect threats independently and accurately, without the need for a human to constantly watch video feeds.

No surprise there — this has been a key part of several megatrends from the Security Industry Association (SIA) for the past few years. But thanks to technology advancements and network convergence, they also have the power to transform how you see your operations with more use cases outside of security. This can include monitoring safety risks on work sites, tracking the quality of the customer experience in retail settings, assessing if equipment needs maintenance in industrial environments, and much more.

Broader applications for video intelligence outside of security have the potential to help people in your organization be more productive, effective and safe in their jobs. However, these applications also come with new considerations for how you and your integrators work together to plan, design and implement video intelligence systems. Understanding these criteria will be vital to unlocking the potential of AI-powered surveillance.

Looking at operations in new ways

Expanding video intelligence to non-security use cases requires comprehensive planning. This proactive approach — versus simply purchasing or upgrading products — helps ensure that your solution best meets your individual needs.

First, integrators need to know what’s happening in your facilities. Through in-depth, design thinking sessions, integrators can help you to explore potential video applications and all the factors that need to be considered to deploy them. These can include the environment, processes, people and their behaviors.

Non-environmental factors also must be considered. For example, anonymized capture may be needed to help prevent bias in an AI system and protect people’s privacy, especially in settings like medical facilities, schools and courthouses. If laws like HIPAA or the EU’s GDPR apply, it can also affect where cameras can go and what data they can collect.

Another element to consider is the offset costs that an AI-powered video intelligence application can deliver. The application may require new investments, but it may also result in new sales or savings that can recoup some or all of the investment. Better marketing and customer experiences that boost sales, better security that improves loss prevention, enhanced safety that reduces injury liability are just some ways costs can be offset.

Considerations for deployment  

More applications for video intelligence beyond security inherently create more deployment considerations. These can range from the camera’s placement and field of view to an environment’s lighting and potential sources of glare to where the analytics will be processed.

Camera placement can be critical to ensure you collect the information you need to measure your targeted analytics accurately. If vehicle activity is monitored, for instance, camera placement should exclude a nearby road in the camera’s field of view that may result in collecting unwanted or inaccurate information. Similarly, an application like people counting or line queueing can be easier to measure if cameras are placed in a straight-down view. 

When it comes to deciding how to process AI-powered analytics, there are four options. Each has its advantages, costs and requirements.

• Analytics at the edge is a good option when using a camera for one application, like fire detection or gauge reading. However, this option can have local data storage limitations and only run a limited number of analytics on a camera at one time.
• Cloud-based analytics is a good choice for applications involving multiple or remote locations. This option is easy to install, scale and maintain. Depending on the use case, that scalability and flexibility can be appealing, especially when compared to solutions such as network video recorders (NVRs) and servers that can be expensive and may not be feasible on a large scale. The trade-off is that cloud-based analytics typically has a recurring fee based on factors like the number of cameras used, the resolution required and how long data is stored. 
• Server-based analytics avoids recurring fees, but an on-premises server must be factored into the overall cost. Manufacturers offer bundles that include cameras, servers and video management systems. If using solutions from multiple manufacturers, make sure they’re compatible.
• A hybrid solution can use a combination of the above approaches to bring each of their unique benefits to a project. For example, you could use on-camera storage and analytics while allowing the device to be managed with a cloud-based application.

Identifying infrastructure needs

The unique demands of each project will determine whether you can use your existing infrastructure or if additional investments are needed.

For example, a line-crossing or area-of-interest application that uses server-based analytics processing could allow you to use your existing cameras. But for applications that run at the edge, you’ll likely need a new, higher-end subset of cameras. 

It’s also important to consider what you want to accomplish long term. Deploying more cameras or cameras with additional sensors could strain network bandwidth, for instance. And cabling limitations could restrict camera deployments to 100 meters from power and connection sources.

Implementing a utility-grade (UTG) cable infrastructure can meet your application and connectivity needs as they inevitably grow in the future. A UTG infrastructure delivers performance and reliability levels beyond industry standards. UTG cables can reach at least 150 meters, some as far as 185 meters.

Innovate to elevate

AI transforms what can be done with video intelligence solutions. It may require you and your integrators to make some disruptive changes to how you design, implement and use cameras. But if you’re willing to navigate this change, you can unlock new, more efficient ways of securing and running your organization.