Artificial intelligence (AI) is reshaping the security industry, driving significant advancements in technologies like video surveillance, access control and video analytics. For systems integrators, the rapid pace of AI adoption presents both opportunities and challenges, from navigating complex implementations to addressing the high cost of these cutting-edge solutions.

The cover story of SDM’s November emagazine, presented as a two-part online series, examines the implications of the rapidly changing AI landscape. The first installment focused on identifying which technologies are most significantly impacted by AI. Insights from interviews with leading integrators offer a closer look at their experiences with AI integration, how it’s reshaping system design, and the ways they are using AI to enhance internal business efficiencies.

This final installment will focus on the practical steps integrators take to ensure effective AI implementation, including assessing client needs, communicating AI’s value and managing budget constraints. Also, perspectives from manufacturers further illustrate how they support integrators in preparing for AI integration through comprehensive training programs and resources.

Effective AI Solution Implementation

When it comes to successfully implementing AI solutions, security integrators must first thoroughly evaluate their clients’ specific needs to determine which AI technologies can offer the most value. From assessing operational challenges to aligning AI with industry-specific use cases, integrators play a crucial role in selecting the right solutions and demonstrating the tangible benefits AI can provide, ensuring that each deployment enhances security and operational efficiency for the end user.

“We start by taking AI out of the picture and instead focus on the problem you want to solve,” says Satish Raj, CTO for San Antonio-based Pro-Vigil, a provider of AI-enabled remote video monitoring solutions. “We’re focused on having a workflow in which we can have accurate crime prevention. We started with that baseline and, once you set your accuracy needs, ask yourself how do you satisfy customer needs at this service level? And then how do you scale this up to thousands of customers? For us, AI is a force multiplier that allows us to offer that quality of service at scale.”

When deciding if AI is a fit for individual customer sites, Pro-Vigil looks at certain criteria including industry type. An outdoor construction site, for example, has different needs than a car dealership. “The same goes for a site that is fenced or gated — what we would recognize as a threat would be different there versus a site that is open to traffic,” Raj says. “From there, we create a customer profile and then we can apply AI to meet specific needs.”

To determine if AI-enabled solutions fit a client’s security needs, it’s crucial to ensure clear benefits like ease of use, streamlined operations and enhanced support, Amir Shechter, executive director of innovation and technology of Convergint, explains. The key challenge is delivering value that justifies the investment, with a strong focus on ROI to drive widespread adoption beyond the proof of concept.

“This is why upfront consulting, advisory services and technical innovation are essential. They help ensure organizational alignment, stakeholder buy-in and a willingness to manage change,” Shechter says. “Establishing realistic goals, achievable technical parameters, a phased approach, and the assignment of the right resources are all key. Ultimately, creating a collaborative project environment that includes internal and external stakeholders — such as the customer, integrator and solution providers — ensures the project’s success.”

Justin Stearns, vice president and partner, Chimera Integrations, also emphasizes the necessity for collaboration. Initially, the technology is vetted in-house before being considered for a client. For larger installations, a demo is set up in the client’s facility to allow the technology to demonstrate its capabilities. For example, Chimera recently designed a 250-camera system for a manufacturing facility. The primary concerns for the solution were employee safety and property damage caused by forklift accidents. A demo camera from a large, well-established manufacturer with previously vetted AI solutions was installed.

“The manufacturer supported us with the demo installation and programming and ultimately the solution failed to detect or prevent the situations we thought it would,” Stearns describes. “Ceiling height in this case caused the analytics/AI not to work without a significant increase in camera resolution that would have made the project go out of budget. We did an about face and installed a second demo with a different camera manufacturer and closed the sale on the project.”

Assessing AI-enabled solutions for a client’s security needs involves understanding their unique environment, pain points and goals, says Priya Serai, CIO, Zeus Fire and Security, which takes a consultative approach to ensure AI adds true value. The first step is evaluating the client’s specific security challenges and desired outcomes, such as real-time incident detection, access control enhancements or operational insights from video data.

The environment and context — including facility layout, foot traffic and existing infrastructure — are crucial for the effectiveness of AI solutions. For example, Serai says, optimizing customer flow in retail spaces involves considering both security and operational analytics to improve sales, staffing or inventory placement.

Scalability and integration are key factors as well, ensuring the AI solution can adapt to evolving needs and integrate with existing systems. Flexibility is essential for long-term value and ease of expansion. Additionally, the client’s readiness for AI, including infrastructure and willingness to adopt a data-driven approach, is considered. AI-enabled security requires continuous learning, refinement, and a shift in how teams use and respond to data, she advises.

“Ultimately, it’s about finding the right balance — the AI solution should enhance security, provide actionable insights, and align with the client’s overall business goals,” Serai says. “At Zeus, our focus is on ensuring the solution not only fits but truly makes an impact on protecting people, property and profits.”

Overcoming Cost & Budget Constraints

Security integrators often encounter financial challenges when implementing AI solutions, primarily due to the high upfront costs involved. However, integrators are finding ways to demonstrate the long-term value and return on investment (ROI) of these systems to their clients. By focusing on the operational efficiencies and enhanced security that AI can deliver over time, they work to balance immediate financial concerns with the broader benefits AI brings to end users.

Many security departments excel in operational management but struggle with the business aspect of security, Shechter explains. Building a strong business case and financial rationale for AI integration can be challenging. It involves clearly defining the specific needs, value and financial advantages that AI offers, as well as creating a detailed cost-benefit analysis.

“Clients may be hesitant to adopt AI due to skepticism from previous digital transformation initiatives that under-delivered,” Shechter says. “Without a solid, realistic plan that assures success, this skepticism can become a significant financial barrier, discouraging investment in AI solutions.”

In the physical security sector, the high initial costs of AI systems can be offset by extending their capabilities beyond security functions, Shechter says. For instance, AI-enabled cameras can also be used to enhance safety protocols, manage logistics and gather occupancy data that aids in more informed decision-making across various operations, adding value beyond security alone.

“AI tools can help manage people’s movement, monitor hazards, enhance employee and guest engagement, and contribute to ESG initiatives,” he adds. “By collaborating with other business units, organizations can maximize ROI and tap into additional budgets, making AI adoption more feasible and aligning with overall business goals.”

By enhancing existing video surveillance infrastructure with cameras that incorporate AI capabilities, many specialized IoT sensors are becoming redundant, streamlining design and maximizing the most use of the camera. This will not negate the need for special sensors in the near term, some use cases are very specific and require purpose-built devices.

Mark Barbaric, senior vice president of technology integration at Prosegur, notes that with the rapid pace of technological advancements, the company is increasingly finding that basic AI functionality is becoming a standard feature in modern hardware and the supporting software. For customers that are on a single video and access control platform, the upfront costs are negligible, he says, while long-term ROI can easily be demonstrated for upgrade to more advanced systems.

“Most of our customers recognize that challenge but have difficulty convincing the C-level to spend limited budgets on major upgrades where technology could rapidly become outdated,” he says. “Our approach is to work with manufacturers to design a hardware solution that is future-proofed and open to various software applications, should the need arise in the future for advanced AI projects that deliver high ROI.”

For Pro-Vigil, building AI models in-house has been a tremendous benefit for the company and its customers; however, it took years of expertise and investment to establish this infrastructure, Raj explains.

“Because of our scale of operation, we process tens of millions of events every month. Therefore we are in a unique position to use outcomes from this processing to train AI on an ongoing basis,” he says. “These are all benefits that are ultimately passed on to the customer in the form of faster, more accurate crime detection and deterrence.”

Balancing the high upfront costs of AI systems with their long-term value and ROI requires a clear strategy, Raj says. This includes implementing approaches that reduce the time it takes for AI-driven benefits to materialize, ensuring that the investment provides measurable returns sooner rather than later. If the benefits are vague and take too long to materialize, it might not be the appropriate solution.

“Having a defined roadmap helps solve the problem faster,” Raj continues. “If you start with a well-defined pain point, you can translate that into the value proposition and ask, ‘Am I willing to pay the upfront cost to solve the pain point?’ For Pro-Vigil, that pain point was the ability to scale. As our company grew, we were not able to scale with the same accuracy and that’s when we brought in AI several years ago.”

According to Serai, Zeus is focused on balancing AI’s high upfront costs with long-term ROI by maintaining transparency from the outset, ensuring clients understand both the initial investment and the potential value over time. The company highlights the likely savings and efficiencies, enabling clients to see the bigger picture.

To ease the financial burden, Zeus often starts with smaller use cases that deliver “quick wins,” showcasing immediate ROI. Scalability is achieved through phased approaches, beginning with pilot projects and expanding as results become evident. Zeus also emphasizes AI’s adaptability, showing clients how systems can improve and evolve, delivering increasing value as AI learns.

“Ultimately, it’s about showing a path from upfront investment to long-term gain while ensuring that each step is aligned with the client’s needs,” Serai says. “We don’t have all the answers yet, but we’re dedicated to finding that balance and making sure our AI solutions deliver growing value over time.”

Stearns notes that while the benefits of AI generally outweigh any additional costs, the biggest resistance typically comes from subscription fees, which clients often question. However, this pushback isn’t directly related to AI being part of the solution itself, he says, but is more of a broader concern regarding ongoing costs. He emphasizes that if the AI solution wasn’t financially viable for the client, Chimera wouldn’t have offered it in the first place.

“We also typically have a good, better and best solution for any technology we are presenting, so if budget is a concern the client can decide what features they aren’t willing to pay for,” Stearns adds.

Stearns challenges the assumption that AI systems always come with higher upfront costs. He explains that Chimera often designs systems using both AI and non-AI solutions, and in their experience, AI-based systems don’t necessarily have greater initial costs compared to alternatives. He emphasizes that this cost differential depends on the specific system and application rather than being a universal truth.

“One thing that saves money on installation is we don’t have to do as much programming for the cameras and typically the UI is easier to use, so we save on end-user training costs,” he explains. “In many cases, we don’t need to install as many cameras when using an AI system versus a traditional VMS that lacks these features. Furthermore, a lot of the VMS companies have HaaS (hardware as a service) models for their products, so now there is very little CapEx expenditure and the system is more budgetable and stays in their OpEx budget.”