People Insights ‘Virtual Door’ for Retail Capacity Monitoring

In Retail, it is critical to always know what’s going on. Malls, stores, or supermarkets need highly accurate insights in capacity statuses and visitor numbers to measure performance and optimize operations. Having the latest data always at your fingertips in real-time also helps to improve decision-making. The pandemic has made capacity monitoring and maintaining limits an even more crucial topic. Of course, while retailers have to ensure that spaces don’t get overcrowded, maintaining the privacy of people is most important. This is why our solution is 100% GDPR compliant.

Why a new capacity monitoring solution is needed? 

Typically, infrared sensors or mobile signals were used to gain capacity insights. However, there are some issues with both methods. For instance, infrared sensors become more and more unreliable in peak times, when multiple people pass the sensor simultaneously. Since it’s impossible to check the accuracy of infrared sensors they deliver black-box results the user has to trust. There are also accuracy issues with using mobile signals since phone locations can vary widely from the actual location of a person. Also, using data from mobile phones can create GDPR compliance issues. Swarm Analytics solves these issues reliably with a 100 % GDPR compliant solution since only anonymous data, but no videos are sent while ensuring high accuracy.

How to get reliable people counting insights with the new Virtual Door? 

Swarm Analytics now offers the virtual door as part of the people insights solution, bringing capacity monitoring to the next level. It’s called the Virtual Door and designed to deliver accurate capacity numbers. People entering and leaving a specific retail space are reliably counted with a smart sensor, based on Swarm Analytics revolutionary AI technology. Since this technology is designed for a three-dimensional environment, it delivers improved people counting results and capacity insights. The technology visually tracks people and recognizes the walked directions like people entering, exiting, or simply passing by. Next to that, the people insights solution offers the possibility to reliably count people in specified zones. Since the data is directly extracted from the video stream, no pictures are sent at any time, which guarantees 100% GDPR compliance.

Benefits of the new Virtual Door solution

  • Trustworthy people counting
  • Reliable capacity insights
  • Accurate counting for doors that are closer or further away
  • Smart data collection with improved deep-learning models
  • Free of charge for everyone with an active Swarm Analytics device license
  • >99% capacity accuracy
  • Still tracks people accurately if they are not visible for a short time
  • Ready to use solution
  • Fast and easy configuration

How to start using the Virtual Door?

First, you need an active Swarm Perception Box license. The device is super-fast to deploy and works with almost any IP Camera via plug & play deployment. Furthermore, it’s easily and centrally manageable in the Swarm Contor Center. In the Contol Center, the Virtual Door can be configured within 5 minutes and a few simple steps. Here is how it works:

  • Click on “Configure Solution”
  • Select the people insights solution
  • Create and adjust the Virtual Door with drag and top (yellow area)
  • Name your new Virtual Door
  • Save, and start receiving data

For a more detailed description, please read the Documentation.

Shopping Center uses the Virtual Door Solution

Our preferred solution partner Systems offers a traffic light system to manage capacity levels based on Swarm Analytics Virtual Door technology. This traffic light system is already in use at a major shopping center, which operates the access control product at all entrances to gain crowd and performance insights and comply with COVID 19 regulations.

An example picture of the access control product of our partner Systems

How to get started

As a partner of Swarm Analytics, you can get started right away and try using the Virtual Door in your Swarm Control Center now. Please contact us if you need any support. If you are not a partner yet, contact

Innsbruck, AustriaPeople Insights ‘Virtual Door’ for Retail Capacity Monitoring

How Can Stores Comply With COVID-19 Safety Rules?

Supermarkets and their employees are working extremely hard to provide fresh food, hygiene products, and essential items to all of us. Even when we stay home as much as possible, supermarkets remain open so we can get our essential supplies. However, preventing over-crowding can be difficult when there are too many people in the supermarket. Making sure that there are not too many people in the store at the same time is one of the major challenges supermarkets are facing. Getting it wrong can result in expensive fines, but controlling capacity manually is expensive.

Everyone, including customers and employees, is concerned about the spreading of the coronavirus. The question is, how can stores easily and cost-effectively comply with the safety rules while keeping their stores open? The employees have more than enough to do at the checkout or in the store, refilling shelves. So, there must be a better way!

The People Counting Solution for Supermarkets and Retailers

Our approach for solving this issue is real-time counting of people, going in and out of the store. This makes it possible to know, in real-time, how many people are inside the supermarket at any time. Our business partners use these counts in their newly developed traffic light system, which shows customers if they have to wait or can enter, using a red and green light.

We want to make it easy for store managers to monitor and manage the capacity statuses of their stores, so customers can be served as well as possible, while employees can focus on their daily tasks. If you are interested in this solution, please contact us directly to discuss your options and to get the system installed at your site.


Dashboards and reports for showing these customer insights (from the in/out counting) are available, too. They could support you, in the long run, to identify changes in visitor numbers, recognize trends, and compare branches.

Innsbruck, AustriaHow Can Stores Comply With COVID-19 Safety Rules?

How Has COVID-19 Changed Vehicle and Pedestrian Traffic?

What happens when people limit their social interactions and work from home? People should have little reason to go outside or use their car, except for going food-shopping or going for a short walk. So, the streets should be quite empty in these times, right? Nevertheless, the current situation can be very challenging for authorities to check and see whether people are actually staying at home. It is not safe to send out people to count traffic and pedestrians or setup new traffic counting devices.

The question is, how can we easily see the impact of Covid-19 on traffic then?

Gathering Traffic Data with Counting Lines

Swarm Analytics allows you to collect reliable traffic data 24/7. The Swarm Perception Box transforms conventional cameras into intelligent visual sensors by “translating” the video stream directly into insights. After you have configured counting lines, using the web-based management platform, you can collect data in real-time about all vehicles and pedestrians crossing this line. Since no image ever leaves the camera, we fully protect the privacy of individuals while providing the essential traffic information to the respective city, town, or authorities. The real-time data can provide the basis for understanding city-wide patterns, identify critical areas, and help you implement control measures more efficiently.

Swarm Control Center – enabling easy configuration of new counting lines
Example: counting line for traffic

Dashboard Examples From Real-World Real-Time Data

To illustrate which traffic data can be gathered with the Swarm Perception Box, here are some real-world examples during the current Covid-19 pandemic. Typically, people would use their car for traveling to work in the morning and back in the evening. In figure 1, it can be seen that traffic peaks disappeared and that overall traffic volume decreased.

Figure 1: Traffic counting data – 7 days
Figure 2: Bus traffic counting data

But what about Public Transport? By filtering the data to show the number of busses, on this road, it can be seen that busses still continue to operate to more or less the same levels (see figure 2).

In theory, with the Covid-19 restrictions, people should be spending more time at home than on the streets. However, when looking at the pedestrian data, there are significant peaks in the afternoons, suggesting that people are going for a stroll in the warmer afternoons (see figure 3). And when people counts are visually compared to defined limits, it becomes easy to identify when and how often the set limits were exceeded (see figure 4). These insights could be used to take the respective safety measures and actions.

Figure 3: People counting data
Figure 4: People counting compared to a defined limit

For more information, visit our urban traffic solution page.

danielHow Has COVID-19 Changed Vehicle and Pedestrian Traffic?

How AI Technology Could Give You More Freedom During COVID-19

Many people have privacy concerns relating to video cameras, especially when this is combined with AI technology. However (when properly controlled and regulated) this could, in fact, be actually something that offers people more freedom as the COVID-19 pandemic progresses. Providing some more freedom will be politically, economically, and socially critical, as it looks like we will need to learn how to live with significantly reduced social contact for the foreseeable future.

We can look at this via an example. Understandably, people are desperate to get out of their homes for some much-needed exercise and fresh air. However, in many countries, exercising or even sitting in a park are now forbidden. Although being in a park on your own carries no COVID-19 spreading risk, these rules need to be in place, as authorities have no easy way to see what is happening in public spaces. Are groups/crowds of people gathering? Are people keeping a safe distance? They simply do not know, so a general rule has to be implemented, for example, all parks are closed. However, it gets worse, when any ‘violation’ has to ‘punished’ as an example to make sure people follow the strict rules, like an elderly person sitting alone on a park bench.

So, if authorities could more easily monitor open spaces in real-time, it means that the rules could be loosened and people could enjoy sitting in the sunshine in the park or going for a short walk, while high-risk grouping or crowding could be automatically detected and addressed as required. Therefore, better monitoring, via video cameras and AI, could, in fact, bring more freedom.

What Swarm Analytics Is Working on to Deliver Meaningful Crowd and Group Insights

In response to the COVID-19 pandemic, Swarm Analytics is working hard to quickly deliver extra AI models to deliver insights into people spacing, groups, and crowds in streets, squares, or shops. The existing capabilities of Swarm Analytics for counting and tracking people will be further developed to also enable group size estimation and density information.

Crowd or Group Size Estimation

To gain comprehensive insights into crowds and gatherings of people in environments, where conventional systems like mobile tracking do not work well, Swarm Analytics group estimation is required. For example at big sports events or smaller gatherings, where it’s very challenging to count visitor numbers or identify group sizes. Additionally, if certain limits of group sizes or visitor numbers are reached, real-time alerts can be used to trigger required actions.

Density Information

For evaluating critical situations or identifying overcrowding in shops, shopping centers, squares, or streets, density insights are required. Are people keeping a safe distance apart? When density levels reach a critical level, automatic real-time alerts can be used to take appropriate safety actions.

danielHow AI Technology Could Give You More Freedom During COVID-19

How Effective are Recommendations for Staying at Home in Times of COVID-19?

Traffic Monitoring and People Insights

Measures for limiting social interactions have recently become more and more restrictive. Nevertheless, we hear news about people meeting in parks to enjoy the nice weather. Also, traffic volumes should have dropped as a result of new guidelines. Could data help in identifying critical areas and taking appropriate actions?


Traffic Monitoring Data

Usually, it is quite busy on the road during the week. Using the AI technology on the Swarm Perception Box, it is now easy to measure traffic by automatically counting all vehicles on a specific road with just a cheap (<€50) IP Camera. Since the data is gathered from a live video stream, you can detect what is happening on the road in real-time and generate immediate alerts. The collected data can also be used to compare traffic volumes for different vehicle types over the last hours, weeks, and months.

In the following real-world example from Austria, over the last few days, it appears that the COVID-19 measures had a significant impact on traffic volumes, causing smaller traffic peaks and lower traffic numbers in general.

Example: Counting line, for capturing traffic and pedestrian volumes

Data for People Insights

If the number of people moving on the streets and in public places is of special interest, people can be counted as well. In the graph below, real-time numbers of people walking on the street are compared to defined safe ‘crowding’ limits. The insights about events where safety limits are exceeded could support decisions about increased safety measures. As the data is collected automatically in real-time, alerts could be sent immediately.

Maintaining the Privacy of Individuals

Although traffic monitoring and data for people insights are generated with the aid of surveillance cameras, the privacy of individuals is fully protected. This is ensured through an innovative approach which “translates” video streams directly in real-time data. No picture ever leaves the camera, but messages, consisting of numbers and words, are sent to the respective operator to put the insights directly into use. Therefore, its also 100 % GDPR-compliant.

danielHow Effective are Recommendations for Staying at Home in Times of COVID-19?