ROC is shifting compute off edge-based biometric readers and onto secure-side GPU infrastructure as part of a broader push toward more intelligent, AI-driven physical security systems, according to Senior Vice President of Commercial Security and Attendance Market Gary Jones in a new interview on the ID Talk podcast.
Speaking on the ID Talk podcast, Jones discussed how the company’s new ROC Access Face 1 platform is designed to move beyond the traditional model of biometric access control readers that only authenticate users standing directly at a door.
“What has changed in recent years is with the massive pace of GPU development, I think customers across the board are more willing to do things on the secure side and leverage that GPU compute technology,” Jones said during the interview. “We can also reduce the footprint, the attack footprint on the non-secure side, bring the size of the product down, bring the risk profile down, but also develop a product that can do a heck of a lot more than just grant access for those few brief moments when somebody stands at a door.”
Jones explained that the shift allows ROC to combine facial recognition with additional AI-driven analytics and workflows, including weapon detection, watch list detection, loitering analysis and crowd monitoring.
“These access control devices are in a prime position in any property,” he said. “They’re positioned to see a lot of what’s going on at an access point, but they’re only ever being used when somebody steps up in front of one of them.”
According to Jones, ROC’s approach is intended to transform access readers into continuously operating intelligent security sensors capable of monitoring multiple streams and events simultaneously. He described scenarios in which the system could identify a known employee requesting access while simultaneously detecting a person on a watch list in the background and triggering customized security workflows in response.
The conversation also touched on ROC’s broader positioning in the biometrics market, including its emphasis on independently validated performance benchmarks and fully in-house AI development. Jones stressed that ROC develops its algorithms internally and gives customers full ownership over deployment environments and data.
“All major vendors have moved their matching algorithms to AI-based algorithms or AI workflows,” Jones said. “Everything is pre-vetted and tested before you get the product. The customer owns the data, customer owns the application.”
Elsewhere in the interview, Jones discussed the practical challenges of deploying AI-powered weapon detection systems, calling the technology computationally intensive and cautioning against unrealistic claims around false positives.
“Anybody in the industry that claims that they’ll give you no false alerts in a weapon detection system is simply not being honest,” he said.
The episode also explores ROC’s recent momentum in the physical security sector, including the company’s ISCS West award win for the Face 1 platform and its broader growth trajectory following its public-market debut earlier this year.
For more insights into ROC’s evolving approach to biometrics, AI-driven analytics and intelligent physical security infrastructure, listen to the full interview via the media player below or wherever you get your podcasts.
–
By the ID Tech Editorial Team
