Video image recognition for the respect of barrier gestures… and not that…
The fact : will image recognition save humanity ?
In France, RATP is using the Datakalab startup‘s solution to find out whether passengers are complying with the obligation to wear masks on public transport. The solution works using image recognition algorithms.
While Amazon is being sued for failing to protect the health of its employees in the midst of the Covid-19 crisis, the Seattle-based giant recently developed a solution to help its employees respect flows in its warehouses. A real “Distance Assistant” for the respect of barrier gestures thanks to screens displaying the flows and surrounding each individual in green as soon as the safety distance is respected. When employees are within 1.8m of each other, they are circled in red on the assistance screens.
The system is autonomous (consisting of a screen, a camera and sensors) as it only requires an electrical connection to operate. It can therefore be easily moved around the warehouse according to the distribution of employees. For Brad Porter VP from Amazon Robotics, it’s simply the same function as speed cameras on the road to warn you that you’re driving too fast.
Thermal cameras are also under consideration, not only in railway stations or airports but also in building lobbies, for example, or shopping centres, as in China. These cameras are capable of knowing the temperature of individuals to within 0.3 degrees. The CNIL has just issued an alert on the use of these cameras with respect to privacy.
These solutions are deployed in several companies and also in stores to enforce this barrier distance between salespeople and customers. It is also a way to measure employee movements and therefore their performance… or even to monitor their health without their knowledge with solutions like binah.ai.
The decoding : Will the management methods be altered by image recognition ?
The democratization of in-store video tracking is well known when it comes to marketing studies on flow analysis at the point of sale. But with the Covid-19, video recognition in stores will certainly take on another dimension for health reasons.
Video analysis could quickly switch to measuring the efficiency of sales forces or operators on logistics sites. These signs have already been identified at NRF 2020 with multiple solutions to monitor the physical position of salespeople in stores to know if they were well placed to best serve customers.
This raises the question of respect for privacy and, more broadly, the relationship of trust between employers and employees.
We discussed the future developments of social scoring in the next 10 years in a previous article. It is therefore possible that we may fall into a kind of professional scoring via image recognition. This would then be a total breach of the contract of trust between the company and the employees, which is reminiscent of the most backward management methods of the decades of the last century.
This is how far we are from the revival of the working methods praised by the confinement with the democratisation of telework and thus the confidence of managers in their teams. Confidence being at the heart of all performance let us hope that this pitfall can be avoided in North America as well as in Europe.