Powering Human Presence Detection with Sensors


The ability to detect human presence can be a pretty powerful attribute for a number of devices. Versions of this technology have been around for quite some time, but artificial intelligence and emerging technologies have bolstered its robustness and use cases. So, just what does this technology do, and how is it successfully accomplished?

Human presence detection or human sensing refers to technology used to determine if a person is present in a particular environment. This type of technology may be used in a range of applications, whether it’s for security or safety, or to enable a device like a smart-home hub to perform key functions.

Challenges with Human Sensing Technology

When human presence detection works accurately, it enables seamless background features and immediate security alerts. The kitchen light turns on once you enter the room and off once you leave. Your personal computer blocks others from using it.

But developing that level of accuracy has its challenges. One of the major challenges is making the distinction between detecting motion versus actual human presence. If the device or system relies solely on motion detection, it can lead to false positives—imagine if your home alarm went off every time a rabbit hopped past your front door. Along with being extra annoying, it would also destroy its real purpose. You’d be so accustomed to false alarms that you’d likely miss a real threat.

This is just one small example of why it’s so important for human presence detection to be as precise as possible. And it can be, with a strong understanding of the context, the right sensor setup, and the help of sensor-fusion software.

Sensors Powering Human Presence Detection

It’s important to know what sensors make up this world. Here’s a list of different sensors involved in human presence detection:

  • Ambient light sensors: As the name suggests, these sensors detect light in the environment. They can measure the light reflecting from an approaching user to detect human presence from a distance. They’re frequently used to automatically activate devices, such as smartphones or computers, without the user needing to actually press a button. Ambient light sensors also measure ambient light intensity to adjust display backlight brightness.
  • Ultrasonic proximity sensors: This is one class of proximity sensor that’s useful in human presence detection. They emit ultrasonic waves and analyze the time it takes to return to determine distance to the sensor. They tend to work well in more extreme environmental conditions (outside of vacuums), but they have a limited detection range. You may have seen these used to detect when a human enters or leaves the vicinity of a device, such as a self-serve kiosk at a mall, airport, or bank.
  • IR proximity sensors: IR proximity sensors work similarly to ultrasonic types, except that they rely on infrared (IR) light to determine distance and detect figures. This is the most common type of proximity sensor used in surveillance and security applications. It can measure the distance to “soft” objects and complex textures, unlike ultrasonic sensors, but its performance can dip over long distances.
  • Capacitive proximity sensors: These sensors work by creating their own electrostatic field. However, when an object approaches the sensor, it changes the sensor’s capacitance, resulting in an amplitude change. This change triggers an output switch. Capacitive sensors can be used to detect boundary penetration through openings such as window frames or ventilation ducts.
  • Time-of-flight sensors: These sensors emit a signal that reflects off a surface and measures the time it takes to return to the sensor. Sound familiar? Well, it’s similar to the IR proximity sensor; however, time-of-flight sensors include multiple types of signals, including lasers, IR, and ultrasound. Time-of-flight sensors use an array of light-sensing pixels to determine relative distances from the object, creating a range map. This allows for a more complex and informative view of what it’s pointing at than the simpler proximity sensor. These tend to be good options for devices requiring low power consumption from the sensor (such as smartphones and consumer electronics).
  • Passive IR detectors: PIR sensors are used to detect motion of IR light-emitting sources (typically human bodies). These types of sensors are all around you; they flip the lights on when you walk into a public restroom and open the automatic door at the grocery store. It doesn’t actively emit IR light, but instead looks for changes in the IR world that it sees. This can lead to false positives for HPD, but it can be reliably used in a context where only humans would be present.
  • Visual sensors: Thanks to image processing and recognition, visual sensors can be used to detect human shapes. These are the sensors that power facial-recognition software across all kinds of applications, including your smartphone and laptop. Emerging artificial-intelligence technologies take this even further: They can recognize and track people based on their specific gait as they walk, along with their heartbeat and microbial traces.

These sensors on their own aren’t enough to accurately detect human presence. Understanding context and how to combine sensors can help make them more accurate. For a clothing store, a shin-level IR sensor might be enough to detect people entering and exiting the store. However, some applications require more than one sensor to detect and gain context from human presence

With a phone, you can use sensors to keep track of the angle at which it’s being held, and pair it with a proximity sensor to ascertain that it’s being held to the face and block unintended button presses. Or, for more advanced functionality, visual or IR cameras backed with facial-recognition software not only can verify human presence, but also track who the users are before allowing access.

Common Applications Using Human Sensing Functionality

As advances are made in AI and other emerging technologies, there’s likely no limit to the applications that can be configured with human presence detection. Two of the most common uses of the technology are security and energy conservation.

In security and surveillance, sensors for human presence detection are often used to detect intruders, whether it’s in your personal home, an office or retail setting, or even an industrial environment. With computers, human presence detection is also used for security purposes. Visual sensors can be used to determine if someone is in front of the computer screen, particularly you.

Human presence detection can also help conserve energy in buildings by powering off lights and other devices if no humans are in the room. In addition, it could be used to determine if you’re still watching Netflix or if you’ve fallen asleep, saving bandwidth and some embarrassment (even if it’s just from yourself).

CEVA is working on algorithms designed to aid in human presence detection for varying applications from earbuds to computers. Contact CEVA-Hillcrest Labs with questions to learn more about achieving human presence detection in your application.

Charles Pao is Senior Marketing Specialist, Sensor Fusion Business Unit, at CEVA.



Source link