Keeping enterprise environments safe and secure with smart devices at the connected intelligent edge

The future of AI-accelerated, feature-rich, intelligently connected cameras is upon us. Qualcomm Technologies camera solutions combine powerful image processing and AI with machine learning (ML). They’re purpose-built to help users create a wide variety of intelligent IoT environments across consumer and enterprise applications, including surveillance, dash cams, smart access controls and panels, and more.

Our connected intelligent camera solutions help to secure and safeguard enterprise environments and are transformative for the enterprise security and public safety segments. Through the Qualcomm Vision Intelligence Platforms and various strong ecosystem collaborations, we help businesses transition from a traditional VMS model to a VsaaS model across a variety of applications, from surveillance to smart cities. Ultimately, this transition helps optimize business operations.

Qualcomm has a reputation for being a pioneer in the mobile and wireless industries, but our IoT business supplying connected cameras is growing steadily. As such, I wanted to take this opportunity to answer some frequently asked questions about that business, especially how it relates to AI, the connected intelligent edge, and security more broadly.

What’s the most important trend in camera technology for the security industry?

Security departments need to see and analyze video and data securely and efficiently to provide meaningful interpretations of a scene. While the initial wave of cameras using computer vision was an important innovation, the advent of ML has enabled the security industry to take a giant leap forward. Right now, there are three technology trends that will continue to drive innovation.

First, cameras are beginning to see more because sensors are rapidly transitioning from Full HD (FHD) to 4K resolutions. Camera makers are also adding multiple sensors to a single camera unit, enabling higher utilization of the underlying hardware.

Second, contemporary cameras have more capabilities than they did in the past. Compute capabilities on cameras are rapidly increasing, and companies are innovating with newer machine learning models in response to the high demand of a variety of use cases, such as people and object detection, classification, people counting, and vehicle recognition.

And third, security cameras have become ubiquitous, making it possible to install cameras in certain hard-to-wire areas. And with 5G becoming more pervasive, the low latency and high bandwidth of a 5G connection enables a variety of new use cases, from mission critical emergency services to smart factories and smart cities.

What problems do advanced cameras solve for the security industry?

The security industry has always had to produce meaningful results with little to no human intervention. This needs to be done in real time without delay, and the privacy of the data must be protected. In response to this challenge, AI workloads must be shifted from the cloud to the edge. Smart cameras have dedicated accelerators and trusted execution environments that run ML-based inferencing on vision data, securely and privately in real time.

Connected intelligent cameras solve many security problems, but they can also present security challenges. What must organizations consider from a network security perspective when deploying IoT cameras protecting critical infrastructure?

Securing the traditional network perimeter is no longer sufficient. With the rise of applications being deployed in multiple clouds and the coincidence of a growing mobile workforce, the traditional network perimeter has disappeared. A Zero Trust approach is imperative to enhance security. It’s important to continuously verify device trust, and monitor device health, of connected cameras to verify the authenticity and integrity of sensitive data used in critical decision-making processes.

Qualcomm Technologies invests heavily in SoC-based hardware and software — as well as trust enabling services — to enable a system-wide approach to securing various use cases. Qualcomm Technologies’ processors are equipped with a host of security features, such as secure boot, full disk encryption, trusted execution environment, secure SPI/I2C protocols, and secure camera interfaces.

These connection security features allow end users — consumers, enterprises, service providers — to establish trust in their devices. And as we expand into services, our cloud service enables end users to establish and maintain device trust throughout its life cycle by leveraging hardware root of trust (HW RoT). Our services also enable secure provisioning of sensitive information, such as ML models from the cloud to the edge.

Upgrading from legacy technology to intelligent cameras can be extremely expensive. How should organizations approach the challenge?

This is a very real problem, one that impacts individual businesses as well as the various OEMs that have invested heavily in deploying intelligent cameras in the last two decades.

The solution is an intelligent edge device — in other words, an AI Appliance or “AI Box” that supports traditional IP camera video streams to be off loaded and processed via the box — powered by SoCs designed for high compute and high concurrency of ML-based analytics across multiple camera streams. For instance, the Qualcomm QS8250 and Qualcomm QRB5165 can decode incoming video streams and run AI models on each stream concurrently, and requires very little power to operate. Such an AI Box enables ML processing at the edge for a fraction of the cost of the cloud. Furthermore, 5G enables more time sensitive and/or bandwidth intensive applications.

The AI Box approach is also aligned with the industry’s shift from video management software (VMS) to Video software as a service (VsaaS) implementation. Deploying AI services to an already operational network of cameras enables businesses to optimize their operations, increase profits, and strategize growth. And they can do this without having to upgrade all their cameras. Eventually, I think we’ll see an intricate web of edge AI Boxes and advanced AI cameras that improve the quality and productivity of businesses and improve efficiency and safety in cities.

Qualcomm is an industry leader in wireless technology. How does that DNA help organizations implement advanced camera solutions?

Qualcomm pioneered several generations of wireless technology as well as breakthrough advancements in camera and AI technology. Building custom SoCs that integrate multiple technologies focused on IoT — powerful ISPs, neural network accelerators, connectivity capabilities, location services, and cellular — is in the DNA of the company. We’re also working with our ecosystem partners to deliver comprehensive end-to-end solutions that allow installers and service providers to offer end-customers and end users with a simple path to adopt these new technologies. It removes the challenges and complexities companies are faced with in deploying new technologies that can ultimately deliver a more secure environment, cost-savings, operational efficiencies, and sustainability.

With each new generation of SoCs, we continue to innovate and provide more powerful AI acceleration at the edge. This makes it easier for our customers to optimize their AI models to take full advantage of the underlying hardware. Our SoCs deliver cutting-edge image quality, and we’ve brought the power of multi-camera to both mid- and high-tier processors. We are therefore constantly addressing the growing demand to capture more and more pixels — and the need to run multiple AI models concurrently in a secure environment.

Does Qualcomm have any new smart camera IoT solutions on the horizon?

Yes! At ISC West, we just unveiled the Qualcomm QC7230 chipset, expanding our Qualcomm Vision Intelligence Platform portfolio. Our latest smart camera IoT solution will support the improvement of security solutions and operation effectiveness with real-time edge computing and AI and ML detection, as well as analytics to deliver safer and more protected cities, communities, and enterprises. The QC7230 will help drive the ongoing digital transformations of enterprise security and the public safety segment — helping to keep enterprise environments safe and secure with smart devices at the connected intelligent edge.

Qualcomm Vision Intelligence Platform, Qualcomm QS8250, Qualcomm QRB5165, and Qualcomm QC7230 are products of Qualcomm Technologies, Inc. and/or its subsidiaries.