AnyVision explores biometrics in food distribution and logistics applications
AnyVision recently hosted a webinar in collaboration with FGF Brands to discuss the potential of face recognition and other biometric technologies in distribution and logistics applications.
During the event, Enrico Montagnino, VP of Global Sales at AnyVision talked to Tom Mansourfar, VP Data Sciences, and Analytics at FGF Brands about the companies’ recent partnership, as well as FGF Brands’ efforts in the field of facial recognition.
For context, FGF Brands uses machine learning and computer vision to automate the high-tech bakeries that supply a number of food retailers, including Starbucks and Walmart.
The “who” in facial recognition
Answering a question from Motagnino, Mansourfar said the deployment of facial recognition in automated factory manufacturing was initially prompted by the necessity of understanding who the people on the work floor were and what they were doing in real-time.
“The ability to add the ‘who’ in operational effectiveness is a huge advantage,” he explained.
In terms of additional biometric applications, Mansourfar explained how the firm used facial recognition cameras to allow employees to clock in and out.
The devices did not feature mask detection capabilities, so employees had to remove their face masks upon approaching the cameras, but they did offer temperature detection.
Health and safety applications
The webinar then proceeded with Mansourfar outlining FGF Brands’ upcoming plans in the field of face recognition.
The executive said the company is currently exploring additional health and safety applications, to identify and direct people, for instance, during fire evacuation procedures.
Furthermore, Mansourfar explained how the pandemic had caused adoption rates of facial recognition technologies within the industry to soar.
This was due to the fact that this type of contactless tools greatly limited the spread of the virus, particularly during the early stages of the pandemic, with FGF Brands recording approximately 96 percent adoption rates in the first few days from deployment.
Tackling privacy and data protection issues
From a technical standpoint, FGF Brands had to enroll thousands of employees in the new face recognition-powered system, taking passport-style photos of each of them.
During the biometric enrolment process, many of them asked how their photo would be used, and Mansourfar said it was important to provide them with a clear answer.
FGF Brands clarified that employees’ photos would only be used for authentication and access to the company’s facilities, and never shared with third-party firms.
The executive reminded the viewers that photos taken as part of enrollment in facial recognition systems are very much like those taken by Human Resources departments to register employees, and they have nothing to do with surveillance purposes.
“That doesn’t add value to the organization, for us to track you,” Mansourfar explained. “We’re not a special government agency, we just want to make your life better.”
The future of computer vision in distribution and logistics
According to Mansourfar, vision (or line of sight) is individuals’ number one sensory input to our brain’s smart decision making, and computer vision is essentially in the process of replacing that.
“Look at your team, look at your operation, look at any decision that’s made using your line of sight, and then think how you would benefit from a computer vision ‘superhuman’ who could make those decisions for you.”
The best part of this shift, Mansourfar believes, is the reliability and accuracy of today’s biometrics and computer vision systems, particularly in relation to their decision-making process.
The webinar then concluded with a Q&A session where Mansourfar discussed the implementation of AnyVision’s APIs within FGF Brands’ systems and the benefits of automation in the distribution and logistics industry.
Article Topics
Anyvision | biometric identification | biometrics | computer vision | data protection | facial recognition | identity verification | machine learning | privacy | time and attendance
Comments