How sensors can ensure safety in public places?

Date: May 27, 2020

Author: Ipshita Biswas

Things are never the same after a humungous disaster takes its toll. The 9/11 attack prompted several high-profile buildings to enhance their security measures. This time it’s the pandemic which doesn’t seem to slow down, and there’s every possibility of us having to live with it. Offices, shopping malls, and restaurants are about to change for good, as every such public place is planning to initiate a range of security and surveillance protocols to curb the spread of the virus.

Alongside substantial architectural changes to guarantee social distancing such as building floors, thermal cameras to frequently keep a check on your temperature, equipment to test your daily health, and devices to keep score of your success at social distancing are about to become mainstream.

According to some researchers, SARS-CoV-2’s rapid spread can be attributed to the movements of people with no or very mild symptoms i.e. those who are unaware of already being infected by the virus. That is why social distancing is such an important containment measure, they explain.

With the mounting concerns around the world about economic health, there is now a growing need to gradually reopen the usual operations. Reopening won’t be that simple. In some cases, reinventing is the only solution to creating a new normal in this unstable and unpredictable situation. In this testing time, prevention is a thousand times preferred over the cure, when the latter still remains a mystery that millions are working to solve.

As a precautionary measure to safely conduct operations in places where it’s imperative for human interaction to happen like hospitals, retail stores, and certain workplaces, social distancing and frequent sanitization become the norm. So, how can it be ensured that the two most vital rules are not violated?

Sensing technologies to ensure social distancing and sanitization

Social Distancing using sensors – PathPartner Sensor Solution

Figure 1- Social Distancing using sensors – PathPartner Sensor Solution

Sensors, which are at the heart of any IoT application, are responsible for amassing data from the surrounding environment. Even with the prime purpose of collecting data from the environment, all sensors are significantly different from one another, and each IoT application, based on several requirements, uses different types of sensors. Proximity, Pressure, Temperature, Bio, Imaging are some of the types commonly used in IoT applications.

Simple tasks like object detection can be performed by most of the available sensors like ultrasonic, photoelectric, and optical sensors. With further integration with several other technologies and algorithms, the detected objects can be differentiated and counted. And that is a common sight in smart factories utilizing automation.

So, going by the same logic, would it not be possible to detect each person entering a premises, automatically dispense sanitizer on them when needed? Further, when the sensor detects more than one person in the same room, won’t it be able to ensure that the minimum distance (6 feet, as per the social distancing norms) is maintained between each individual?

This is definitely a possibility, but there do exist a few challenges when it comes to the detection of human presence.

Challenges in people detection and counting

The following are a few challenges while detecting human beings –

  1. How to differentiate between a human being and inanimate objects like chairs, tables, etc.?
  2. How to account for the different heights of different individuals?
  3. What if it’s a very large room with several inanimate objects?
  4. What if the room isn’t always illuminated or is smoggy?
  5. What if I have a large premise? Won’t I be required to install too many sensors?

mmWave Radar technology to the rescue

Millimeter-wave (mmWave) is a special class of radar technology that uses electromagnetic waves of millimeter range (i.e. a short-wavelength) for the detection of objects and providing the range, velocity and angle of these objects. It is a contactless-technology that operates in the spectrum between 30 GHz - 300 GHz.  The use of small wavelengths has certain advantages when it comes to accuracy, penetrability through certain materials, imperviousness, and infrastructural requirements.

The above-mentioned characteristics of mmWave Radar sensors help it overcome most of the challenges associated with people detection and counting.

Automatic Sanitization – PathPartner Sensor Solution

Figure 2- Automatic Sanitization – PathPartner Sensor Solution

Detection of human beings – Due to its accuracy, it can detect even small bodily (respiratory movements) movements, thus differentiating living objects from the surrounding inanimate objects like chairs, tables, etc. Due to its penetrability through certain materials like plastic, drywall and clothing, its readings aren’t impacted by the presence of any such materials around.

Due to the use of short-wavelength, the size of system components, such as the antenna required to process mmWave signals, is small. High resolution being another advantage of short wavelengths.

Imperviousness to environmental conditions – mmWave technology isn’t impacted by the presence of rain, fog, dust, or snow and works perfectly fine even in a murky room.

Based on TI’s mmWave Sensor, we have devised a ready-to-integrate SDK that is apt for detecting and counting people, overcoming the remaining challenges. Our solution also supports NXP, Infeneon radar sensors.

PathPartner’s solution

- PathPartner Radar SDK for UV LEDs

 Figure 3- PathPartner Radar SDK for UV LEDs

The given diagram demonstrates people-counting demo pipeline running on TI’s mmWave Radar sensor using DSP C74x and ARM - Cortex R4 core.

CFAR (Constant false alarm rate) adaptive algorithm is used to single out the signal returned from the target against the noise and clutter in the background.

The Direction of Arrival (DoA) estimation of the detected object is done by MUSIC algorithm. We can provide a wider FOV, thus enabling the coverage of a large area with lesser sensors. Up to +/- 60 degrees, our accuracy remains well within 1 degree. Thus, people of every height will be detected accurately. To date, our best accuracy achieved is 0.1 degrees.

Gtrack is the main algorithm behind object tracking.

PathPartner Radar SDK for UV LEDs

 Figure 4 – PathPartner Radar SDK for UV LEDs

The given diagram demonstrates the workflow of our people-counting solution.

The Radar SDK will be connected to a PC via USB/UART.

A PC can do the initial configuration of the Radar SDK and trigger the sensor.

Radar SDK will then capture the real-time data and process the data on internal cores.

The output of the processing will be displayed in Radar Application GUI on a PC.

The number of detected people will be displayed in the GUI accordingly.

Further Reading

Want to know more about our Radar-based people-counting application? Leave us a mail at

By submitting this form, you authorize PathPartner to contact you with further information about our relevant content, products and services. You may unsubscribe any time. We are committed to your privacy. For more details, refer our Privacy Policy

Camera & IoT

By submitting this form, you authorize PathPartner to contact you with further information about our relevant content, products and services. You may unsubscribe any time. We are committed to your privacy. For more details, refer our Privacy Policy

Notify of
Inline Feedbacks
View all comments
Back to Top