By Srinath Kalluri, CEO, Oyla Inc.
In recent years, there have been groundbreaking innovations made in video surveillance technology, most notably around the application of video analytics and artificial intelligence (AI) solutions. What’s clear is that AI has real staying power, as the market itself is being driven forward by an increasing focus on integrating advanced machine learning algorithms and intelligent analytics.
AI is finally beginning to be more widely adopted, and the awareness of this practical next-gen technology is in a full snowball effect. Leaders in the security industry have taken notice of the fact that AI-driven video analytics are finally becoming robust enough to be widely deployed in the field, driven by enhanced computing capabilities, an increasing array of low-cost sensors, edge-computing platforms, and cloud-enabled architectures.
There’s just one glaring issue: this technology has been plagued with an inability to deliver true, real-time, actionable intelligence. Companies are now thinking more strategically based on recent lessons learned. Businesses as a whole are converging security and operational technologies and shifting to a more proactive and predictive model as opposed to being generally reactive.
The Oyla solution encompasses this ability to offer real-time intelligence by fusing a new piece to the AI-driven puzzle: 3D data. Imagine leveraging 3D data to enhance existing video surveillance technology, and applying AI driven detection algorithms. Oyla uniquely integrates these technologies at the hardware level, and feeds the data to an advanced AI platform.
Dynamic and intelligent LIDAR (Light, Detection, and Ranging) technology generates 3D visibility of any physical space. But, what differentiates Oyla from other manufacturers is a 3D sensing component that is not only more affordable but also integrated in a way that inherently fuses the 3D and high resolution video data together.
The result? An end-to-end 3D-aware surveillance system that integrates LIDAR with video to significantly improve the accuracy of video analytics. We call that better data for better AI.
It’s a matter of trading simple 2D rules-based approaches in video technology for more deep learning-enabled functionalities, which better meets end-user expectations in the market today. LIDAR provides a rich set of data about a scene, so combining that information with images on the visible spectrum to identify, measure, and track people and objects in real-time makes this fusion of technology into a single solution not just a no-brainer, but the future of intelligent video surveillance.