Seeing Around Corners: Video Analytics in Real Estate

Pruven Capital
Pruven Capital
Published in
4 min readSep 6, 2022

--

By: Travis Skelly, Partner at PruVen Capital & Jared Katzen, Investor at PruVen Capital

With violent crime rates increasing by nearly 50% between 2019 to 2021, renters are citing security as their top priority when considering where to live [1]. This shift, plus the growing concerns around gun violence, has created a pressing need for new security solutions that proactively deter, identify and intervene during criminal activity. To address these concerns, property owners have begun experimenting with a variety of new solutions from autonomous security robots, connected access control solutions and, most interesting of all, computer vision video analytics that make security cameras “smarter”.

Market Landscape: Having spent several months diving deep into the space, including dozens of calls with companies throughout the ecosystem, we’ve developed a market map highlighting some of the leading players in the space. Through our analysis, we determined three models into which companies can be classified:

  1. Open Ecosystems, which allow for camera interoperability and flexibility and rely on a separate video management system to connect independent camera nodes
  2. Closed Ecosystems, where the cameras are made by the same manufacturer and are independently internet enabled with edge computing capability that connects directly to the cloud, and
  3. Niche Solutions that serve a more specialized role in the market, acting as more cost-effective point solutions for specific use cases like gun detection and access control, or serving unique verticals such as restaurant & retail industries:

Computer vision enhanced security solutions supercharge existing camera installations, allowing security professionals to focus on responding to and triaging incidents, as opposed to scanning dozens of screens for suspicious activity. Protecting physical assets used to revolve around a few simple principles: 1) limiting the number of access points by locking doors and windows, and installing fences and gates, and 2) monitoring access points closely with a combination of cameras and security guards. With enough resources, these traditional methods are effective, however, there are obvious physical and economic limitations. For example, security guards can’t be present in two places at once, they aren’t available 24/7, and they run the risk of overlooking an issue due to human fallibility. Security cameras with computer vision can solve these challenges, allowing for rapid response and intermediation for events like: vehicle identification, facial authentication, remote perimeter sensing, firearm detection, and injury and fall detection. However, not all properties find the cost of implementing these camera installations and systems within their operational budget as systems often require a minimum number of cameras in order to be effective and, in some cases, require edge computing solutions that can become prohibitively costly with smaller installations, which is why we at PruVen are most excited about systems built for Open Ecosystems.

Why Now

AI powered video analytics tools have benefitted from three key technology enhancements that have yielded a dramatic improvement in performance. First and foremost, improvements in camera quality with edge computing capabilities as new features that enhance video capture have been introduced like HDR, low-light infrared and dewarp technology, which is enabled by specialized chipsets from manufacturers like Ambarella designed for image capture and processing on the device at the edge. Cloud computing cost for video processing also continues to diminish as video database design improves, making it easier to organize and store video clips for event recap and recording. Most impactful of all, there’s been a dramatic improvement in AI/ML techniques in computer vision, which identifies visual objects of certain classes and provides context to perform operations, thanks to advances in object detection and Deep Learning techniques that allow for an ever accelerating learning process using increasingly complex training sets.

AI enhanced video analytics solutions are quickly becoming the gold standard when it comes to monitoring physical spaces. Coupled with the right operational strategies to intervene when situations arise, these technologies can significantly reduce both instances of crime as well as the time it takes to apprehend perpetrators when crime occurs, creating safer spaces for all. If you’re building security solutions for real estate owners we would love to connect. Please reach out to PruVen!

Source: National Commission on Covid-19 and Criminal Justice

--

--