Proximity Sensor Research

Group Members: Jean, Melanie

Sensor Description

Adafruit APDS9960 as described on its website, is a compact, feature-packed sensor breakout board that integrates proximity sensing, ambient light sensing, RGB color detection, and gesture recognition into a single module. It can detect simple gestures (left to right, right to left, up to down, down to up are currently supported), return the amount of red, blue, green, and clear light, or return how close an object is to the front of the sensor. Designed for ease of use, this sensor connects to microcontrollers via an I2C interface, making it simple to wire up and integrate into projects. 

The sensor from Avago Technologies uses an integrated IR LED and four directional photodiodes to detect reflected IR energy. These photodiodes are strategically placed to capture changes in light reflectance from different directions, allowing the sensor to interpret gestures and measure proximity. For RGB and ambient light sensing, the sensor analyzes the intensity of red, green, blue, and clear light using its built-in color detection system.

Here’s a video of the sensor in action:

 

Detection Attributes

Proximity Detection

Sensing angle is approximately ±30°.

Gesture Detection

The sensing angle is wider to capture the movements, but the detection is based on the position change relative to the 90° directly above.

RGB and Light Detection

The sensor has a wide sensing angle of about almost all ranges of ±90°, but the data change unstably and will change depending on the condition of the light.

RGB and Light Detection

The sensor has a wide sensing angle of about almost all ranges of ±90°, but the data change unstably and will change depending on the condition of the light.

Min/Max Values

Gesture Detection

Effective range for gesture detection is about 1-20cm, and the detection accuracy is best at the distance of 5-10 cm. Movements can be detected in areas that are closer (1-5 cm) or further away (15-20), but accuracy of detection drops dramatically. 

Here’s a video of being too far from the sensor:

 

Here’s a video of being too close to the sensor:

 
 

RGB and Light Detection

There is no specific distance limitation, but it appears the proximity of an object can affect the range and intensity of light that can be detected, resulting in inaccurate detection.

 

Proximity Detection

Max distance the sensor can detect is about 20 cm, and any distance farther than 20 will send a number 0. Minimum distance is about 2-3 cm, and the value stays at 255 even as I move closer from 2 cm.

 

Strength/Weakness

  • The sensor is compact and integrates multiple sensing capabilities in one module (gesture, proximity, ambient light, and RGB color sensing), which make it easier if multiple combined sensing are required, and in an energy-efficient way. 
  • The function of detecting hand gestures without physical contact can be an interesting and useful intangible interaction to use. In the situations of slow simple movements, the sensing is relatively accurate. However, if the movement is fast and multiple repetition of movements or complex gestures will make it difficult to detect. 
  • As the gesture detection and proximity are infrared-based, the sensor works well even in low-light conditions. But for the color and light detection, the sensor is not accurate and is too sensitive to the change in the light intensity. 
  • The detection range and sensing angle could be enough for small-scale projects, but it would be too narrow for bigger scale projects.

Existing Application/Inspiration 

Meta is currently developing hand gesture recognition watch to replace mouse, it is possible that it requires proximity sensors to detect gestures like swiping right, left, up or down.

Smartphones black-screening when held near the ear is also an application of such sensors. Also, some clocking-in mechanisms through actions can be used with simple gesture sensing through the sensor. 

Leave a Reply

Your email address will not be published. Required fields are marked *