Smart Building IoT by 20303
By Tara Dunning , Wesco; Kasia Hanson, Intel | October 23, 2023
Read Time: 12 Minutes
As sensors and analytics expand the use cases for security cameras, integrators must keep several factors in mind as they design new or upgraded systems for customers.
Since its introduction in the 1940s, the surveillance camera has been an iconic and central component of almost every security system. For most of its lifespan, integrating the camera into a security system was fairly straightforward – most applications simply needed to have it powered, connected, and positioned to have a clear view of the desired location.
But the camera is now much more than a security device. The rapid growth in advanced video analytics, powered by artificial intelligence (AI) and deep learning, has transformed the surveillance camera into a powerful sensor. This has greatly expanded the use cases for cameras, and as a result, increased the design and specification considerations. When designing a system, integrators must now consider not only the intended application, but also the desired analytics, data and processing capabilities, and existing infrastructure, among others.
In a recent interview with thought leaders at Intel and Wesco, we explored the implications of “camera as a sensor” on surveillance system design. Kasia Hanson, Intel’s Global Director of Security Sales, and Tara Dunning, Wesco’s Vice President of Global Security, discussed emerging trends and the new considerations integrators must consider with new use cases.
The evolution of the camera and its capabilities means that both integrators and end users must consider several criteria that can significantly impact system design.
The desired application will potentially have the greatest impact on overall surveillance system design. Camera placement, quantity, type, lighting, network infrastructure, processing requirements, and edge versus cloud solutions will all be dictated by the challenge the system is trying to address. For example, if the primary goal is to monitor line queuing, the camera will need to be placed with a straight-down view of the area in question. Incorrect placement can have a negative impact on system accuracy. Similarly, license plate recognition systems require specific camera angles and resolution, which must be properly incorporated into the overall design.
The application will also determine the number of cameras. If the goal is loss prevention, for example, it’s critical to ensure that the design incorporates enough cameras to adequately monitor the designated area, whether that’s self-checkout, product shelves, or cashier kiosks.
Specific types of cameras might be required for certain applications. If the system also needs to include sound recording capabilities – such as a gunshot detection system for law enforcement – the cameras would need to have built-in microphones.
Certain applications will also require sufficient lighting. Surveillance systems that need to detect or capture clothing or vehicle color, may need to incorporate white lights or have additional lighting installed.
The bottom line is that integrators need to account for the intended application that may be outside the traditional security installation when designing a surveillance system. If it’s not adequately considered in the early stages of system design, the outcome might not meet expectations.
Your use case will have a significant impact on the overall system design. Certain use cases lend themselves to easy deployment within existing cameras. For instance, analytics such as line crossing, area of interest, heat mapping, and people tracking may only require slight adjustments to the existing camera angle or field of view.
However, other use cases, such as line queuing mentioned earlier, may not be able to utilize an existing camera, depending on placement. Additionally, it’s also important to consider whether moving a camera might produce better or more accurate results. An existing camera technically might be able to capture the desired analytics but moving it to a better angle could help avoid data anomalies or reduce noise and false positives. For example, a camera that’s near a window could falsely indicate a reflection as an occupant. Moving the camera so that it doesn’t face the window could alleviate the issue.
Integrators will also need to consider the deployment method. There are four main types of deployments: server-based, on the cloud, edge, and hybrid. Each of these has different costs, advantages, and requirements that may factor into system design.
Not surprisingly, server-based analytic processing will require some kind of on-site or on-premise server, which will need to be incorporated into the overall cost. While many manufacturers offer bundles that include cameras, servers, and video management systems (VMS), integrators and end users will also want to make sure that any solutions are compatible when using multiple manufacturers.
This option eliminates the need to have significant on-premise server and storage capacity, but generally includes a recurring fee based on the number of cameras, resolution required, and length of time the data is stored. Cloud options can be highly beneficial for deployments that include multiple or remote locations. While cloud-based analytics are generally easy to install, scale, and maintain, it’s also important to consider whether existing bandwidth will be sufficient. While rare, cloud systems can also go down. If 24/7, on-demand access is critical, additional backup systems may also be needed.
Analytics at the camera – or on the edge – utilize cameras that can load specific analytics directly onto the camera. Processing analytics at the edge typically has reduced bandwidth and server processor requirements. This can be a good option when a camera is deployed for one specific application, such as gauge reading, barcode reading, or fire detection. Edge-based analytics also promote greater resiliency and redundancy. If a network failure occurs, the camera can still capture, store, and support operations locally, whereas a solution running entirely in the cloud would cease to operate. However, it’s important to understand that there may be local data storage limitations, as well as limitations on the number of analytics that can run on a camera at one time.
A combination of on-premises, on-camera (edge), and cloud-based services can offer a great deal of flexibility and offer benefits for each method of deployment. For instance, you can have on-camera storage and analytics while managing the device through a cloud application. You can store a period of footage on-site, and back up older recordings to the cloud. Perhaps select analytics could run on the camera, while deeper analytics that require more powerful processing are handled by a cloud application.
In general, many camera deployments can work within an existing infrastructure, provided it has access to both power and network connectivity. But integrators today may also have to contend with bandwidth and cabling distance needs that go beyond current standards and that current infrastructure can’t achieve.
Incorporating more cameras, or cameras with additional sensors, can strain network bandwidth, and this issue will only grow as more cameras get added. Additionally, cabling limitations mean that cameras can be a maximum of 100 meters from a power and connection source.
However, by implementing a Utility Grade Infrastructure (UTG) cable, you ensure increased bandwidth and extended distance are available, meet your current needs, and are well-positioned to meet future needs as well. All UTG cables can reach at least 150 meters with up to 30 watts of PoE power, with some going as far as 185 meters when used with UTG-rated cameras, all with third-party verification.
Apart from the cabling infrastructure, if your use case required additional lighting, integrators would need to ensure that the camera’s location could also accommodate that. Or, if the application required the camera to be moved, the integrator would need to make sure that the new location could support the device. This could affect installation costs.
End users and integrators may also need to incorporate other considerations, depending on the site’s location and industry.
Medical facilities, schools, courthouses, and many other buildings require anonymity in their surveillance data capture. In those settings, it’s critical to ensure that any cameras and analytics solutions do not capture or store data that could be used to identify an individual or group. Many analytics and cameras have privacy settings that blur certain areas of the image.
Along those lines, HIPAA, GDPR, and other similar legislation place restrictions on recording personal information. Additionally, some laws prohibit recording in areas where there is a reasonable expectation of privacy, such as in an employee break room. Different national, state, or local governments may also restrict data gathering or disclosure requirements. The key takeaway is to have a firm understanding of the local laws at the installation site, since that can influence where cameras can go and what data they can collect.
Implementing a new surveillance camera solution can be a significant investment, but it’s important to also consider how the system can offset other costs or improve efficiencies in other areas. Improving efforts to improve loss prevention or enhance personal safety can lower injury liability or lead to long-term costs savings. Reducing the number of false alarms and making it easier to identify or search for incidents can also help improve productivity. The data provided by these analytics could also provide outreaching cost savings. For example, people-counting and heat mapping data could help enable more efficient use of labor as well as more effective merchandising.
In addition to the considerations of today, integrators and end users should also keep an eye on several exciting and transformative trends in the industry. Having these top-of-mind will help businesses fully leverage the power of security cameras and protect their investment.
It’s almost impossible to talk about surveillance cameras today without talking about artificial intelligence (AI) and machine learning. Like many other aspects of society, AI has radically transformed the capabilities and possibilities of security cameras, and its importance will only grow in the coming years.
“AI is advancing how we use cameras and how they are critical sensors,” Kasia Hanson said. “Eventually we’re going to get to a point where they’re collecting data and that data is being used instantaneously.”
As generative AI continues to evolve, it will help to advance threat assessment, detection, and response. This grows the camera into new, innovative use cases. According to Kasia, for example, municipalities are using advanced cameras to help improve road safety by rapidly identifying traffic incidents and directing response crews so they get resolved faster.
Ethics and guidelines around AI and data usage will also become a larger part of the conversation – especially when it comes to surveillance cameras and security. While governments are developing laws and regulations around AI, companies continue to lead the charge.
“Larger companies are already instituting guidelines and processes to ensure not only AI trustworthiness but AI ethics,” Kasia said. “It’s really important for companies to install their own policies, procedures, and guidance, and not wait for government policies.”
Building your own guidelines for the safe and ethical use of data can do more than reduce risk. According to Tara Dunning, Vice President of Global Security at Wesco, doing so can create a competitive advantage.
Establishing a framework protects your business, protects your customers, and facilitates the adoption of new technologies by streamlining decision-making. That transparency and clarity can provide peace of mind – both internally and to customers.Tara Dunning
Physical security and cybersecurity used to be distinctly separate arenas. That’s no longer the case, and sensor-enabled cameras will highlight the convergence of the two fields and the importance of a robust cybersecurity strategy.
Increasing the number of connected cameras will increase the number of potential vulnerabilities and opportunities for cyber attackers. End users will want to ensure that all those devices are protected, and integrators will need to become proficient in cybersecurity. Vendors that don’t adopt cybersecurity as a key part of their strategy run the risk of becoming irrelevant.
To underscore that point, Kasia highlighted just how vigilant companies need to be.
There have been 800,000 cyber-attacks this year, and those are just the ones we know about. That’s more than 2,000 attacks per day – and that includes attacks on connected cameras.Kasia Hanson
Too many of those attacks are successful. Tara noted how there were more than 1,800 data compromises in the U.S. in 2022.
“According to a recent report from Statista and CNET, more than 420 million people were impacted by data breaches in 2022,” she said. “Data security absolutely must be part of the conversation with integrators and end users.”
On a related note, the growing footprint of connected cameras also increases the need for remote management, maintenance, and firmware updates. Both integrators and end users will need to keep this in mind when upgrading or designing new systems.
As camera capabilities and use cases expand, education – on both the integrator and end-user side – will become critical. Integrators who understand how the technology functions, how to maximize it, and how to train customers on it will become valuable partners and advisors. That may even open opportunities for new revenue streams. However, to leverage these opportunities, integrators may need to develop new skill sets and expertise.
On the end-user side, companies that think outside the box may find opportunities to use camera and sensor technology in applications beyond traditional security. For example, video analytics can also play a significant role in an organization’s safety management program by recognizing PPE compliance, identifying unsafe motion, or detecting near misses in warehouses, among other applications. This could help justify investing in new camera solutions and could lead to additional opportunities to improve business outcomes.
The combination of cameras and sensors has greatly expanded the role that surveillance cameras play in an overall security system. For integrators, the variety of innovative use cases means that designing a system is no longer as simple as mounting a camera and having it point in the right direction. They must now account for a variety of other factors, such as the desired analytics, existing infrastructure, and local laws.
The upside is that this will open new opportunities for both inside and outside the security space. However, meeting your customers’ goals and objectives will require an ecosystem of partners with the expertise and reach to help deliver world-class solutions.
Wesco and Intel bring unique perspectives and distinct capabilities to foster a global ecosystem that helps clients fully capitalize on the power of digital technology for business transformation. With unmatched capabilities at scale, Wesco and Intel together deliver world-class solutions, from edge to cloud.
ABOUT THE AUTHOR
Tara Dunning, Vice President of Global Security, Wesco
Tara Dunning joined Wesco in 2021 as Vice President of Global Security, focused on driving advancement, analyzing go-to-market approaches and rising to meet emerging trends across segments, verticals and geographies. Dunning has had a successful career delivering transformative growth in the IT and OT industries, holding multiple sales and strategy positions across global distribution, manufacturer and service provider companies.
ABOUT THE AUTHOR
Global Director, Security Ecosystem Strategy and Partnerships, Intel Corporation
Kasia Hanson leads Intel’s global video, safety, and security sales for the company’s Network and Edge Group. Her team drives technology design, go-to-market, and ecosystem orchestration strategies that drive revenue and adoption of artificial intelligence and edge technologies across the security industry. Hanson is a 22-year veteran of Intel with deep technology expertise across the Internet of Things, AI, physical security and cybersecurity, data center, and storage.
Submit your information and one of our experts will be in touch shortly.