Yunchao Gong, VP of Engineering, Cameras and Jake Leichtling, Director of Product Management, Access Control at Verkada have shared their thoughts on the developments we might expect to see during 2025, in the fields of AI in Security, Cameras, and Access Control.
Firstly Yunchao Gong makes his predictions on:
AI in security
● Large language and vision models reduce investigation times and unlock deeper insights. AI is already transforming the ways that security teams search for people, vehicles, and objects, and that will only continue to evolve. Soon, organisations will be able to use large language and vision models to search through a wider range of situations, like using natural language to search for specific sounds or actions like a window breaking. Ultimately, these new efficiencies will save precious time in moments when—especially for security and safety—every second matters.
Cameras
● Convolutional neural network (CNN) based computer vision at the edge will drive faster, more accurate real-time monitoring and enhanced decision-making capabilities in security. For example, while legacy systems might rely on basic motion detection or predefined triggers, a CNN-powered camera can instantly differentiate between an authorised employee and an unauthorised intruder, alert security teams, and trigger automated responses such as locking nearby doors or redirecting cameras for better coverage. Ultimately, this will drive more innovation around how security operations are managed and optimised.
● Edge-based transformers unlock deeper insights with greater accuracy. Edge-based transformers excel in accurately identifying precise motions or activities, which will be a game-changer when it comes to safety and security. Take hospitals, for example—monitoring patients for falls and ensuring immediate response is a constant priority. Compared to traditional transformers that rely on sending data to the cloud for processing (which can introduce delays and errors), edge-based transformers process everything locally on the device. This enables more accurate fall detection and real-time alerts, ensuring faster and more reliable responses when it matters most.
Jake Leichtling discusses the Access Control segment and how that might change in the near future:
Access Control
● Mobile credentials will become ubiquitous. Smartphones have become an essential multifaceted tool for everyday life and, as a result, an item most people wouldn’t think of leaving behind. This has primed consumers to adopt mobile credentials. Because they already rely on their phones for everyday life, people want to access mobile credentials across more parts of their lives—from digital driver’s licenses in place of physical identification at airports to mobile key cards for hotel rooms. Digital wallets like Apple Wallet are also making it easier to deploy mobile credentials at scale. This appetite for convenient, secure identity verification and access control solutions is just at the beginning and will likely drive mobile credential adoption across nearly every other use case and sector.
