In these unprecedented times we're using AI to help businesses adapt
We have seen the impact of Covid-19 on our retail customers and recognise the importance of adapting in the current climate. In order for us to best serve our clients and the wider Retail Industry, we have worked with essential service providers to understand the real world challenges of maintaining social distancing in stores to keep customers and colleagues safe. Our CV19 occupancy solution has been developed using our existing technology to achieve this whilst protecting retail revenue.
Challenges facing retail
Inaccurate and inconsistent implementation of social distancing regulation could potentially put customers at risk if store occupancy exceeds guideline levels.
We estimate that on average 2FTE are required per store in order to manually monitor and regulate the flow of customers to keep store occupancy within safe limits.
Trials conducted with other essential retailers show that manually counting customers tends to underestimate the number of customers leaving the store. This leads to stores becoming underpopulated and thus lost revenue.
Using in-store staff to regulate the flow of customers to maintain occupancy levels in-line with government social distancing guidelines could cost a retailer with 500 locations up to £20,000,000.
- We have developed a highly accurate (99.5%+*) software solution using sophisticated, resource-light, deep-learning and machine vision techniques to automatically count customers in and out of stores.
- This solution is applied to existing on-site entrance/exit camera feeds and the data processing then automatically operates a “traffic light” system placed at the entrance to tell customers when they can and can’t enter the store.
Existing CCTV Technology
Our solution integrates with your existing CCTV infrastructure to minimise installation costs and increase the speed and ease of deployment
Machine Vision Module
Our lightweight deep learning software can be deployed on your existing compute infrastructure or using our stand-alone edge compute module.
We can work with you to design a customer messaging and feedback system that fits your brand and communication style.
*accuracy achieved over 1-hour period of continuous operation in a live environment using vertical CCTV cameras.
Our Solution (lite)
- A software-only variant of our occupancy solution that monitors customer numbers against the store maximum and requires no physical installation or additional hardware.
- This solution applies our sophisticated machine vision technology on existing infrastructure and CCTV cameras and uses email or the store address system to alert store managers when the store is approaching capacity.
No Physical Installation
Software variant avoids the need for any physical installation, decreasing roll out time and saving money.
Live system keeps store managers informed of occupancy at all times, allowing them to deploy a staff member to the doors when appropriate.
Existing Store Address System
A demonstration of our machine vision system running against the the Oxford Town Centre Dataset.
Original Data: Benfold, Ben and Reid, Ian. “Stable Multi-Target Tracking in Real-Time Surveillance Video”. CVPR 2011. Pages 3457-3464.
To talk to us about keeping your colleagues and customers safe whilst also protecting revenues, contact us using the form below or email us at firstname.lastname@example.org