Skip to main content
All CollectionsAir Quality SensorsSensor Data
Best Practices for Vape Detection
Best Practices for Vape Detection

Learn how to accurately detect vaping activity

Updated over a week ago

With a collection of up to 15 embedded sensors, Verkada air quality sensors simultaneously measure air quality, temperature, humidity, motion, noise, and more. School leaders can configure the device to display which sensor data they want to monitor, and set custom alerts for when certain thresholds are exceeded. Users can then receive alert notifications in real–time, allowing for fast and proactive responses.

How it works

As vaping occurs, users can see spikes in Verkada’s Vape Index, which measures against a range of different onboard sensors to identify the likelihood of vaping or smoking occurring on a scale of 1 to 100. Administrators can set custom thresholds and time delayed thresholds to receive real-time alerts via SMS and email (or webhooks via our API). When detected, an alert can also be sent to pre-selected faculty members and on-campus security, allowing teams to respond quickly.

Vape detection algorithm

The new, improved vape detection algorithm displays values in a more binary fashion, with a higher confidence level when making a decision about whether an event is a true positive, if thresholds for certain indices are successfully met.

Custom threshold values

Changes in the event detection and analysis phases now account for how the triggers are set for custom thresholds values, which includes the time the sensor needs to make an accurate prediction to flag an event as a positive vape event.

  • A higher sensitivity generates a higher volume of events as the analysis time needed to create an event is lower.

  • A lower sensitivity causes the sensor to gather environmental data for a longer duration to make a more accurate prediction (low being the default).

Custom triggers

When you set a custom trigger now, it now allows you to choose the values at which the sensor now creates an alert. By default, the trigger is set to 50. For most deployments, this should be a good starting point. If you notice that you get alerts for false positives at around 50, then consider adjusting the trigger; for example, if you're getting false positives at around 60, and true positives at around 70. In this case, we recommend to adjust the trigger to be at around 70.

Sensor data

All sensor data is accessible from the Verkada Command platform and is set against a timeline to give administrators a complete view of when these events took place. Additionally, administrators can pair a Verkada camera with any sensor, providing a layer of visual evidence to see exactly what happened. For locations, such as bathrooms, cameras can be placed outside these private areas, giving administrators a non-invasive way to monitor these incidents.

Installation requirements

If installed for vape detection, you should install the air quality sensor on ceilings. Wall mounting is possible, but degrades performance for accurate vape detection and events.

Recommendation. We recommend installing at 8 ft (2.4m) and a maximum height of 9 ft (2.7m) for best air quality measurements. If air quality is not an intended use case, the device can be mounted at other heights or orientations.

Range of air quality sensors for vape detection

When mounted on an 8 ft (2.4m) ceiling, the SV11, SV23, and SV25 can detect vaping events within a 250 sq ft (23 m2) area around the sensor, or a 8 ft (2.4m) radius. Higher ceilings or increased distance from the sensor decreases detection accuracy, as the vape/smoke particles are less likely to reach the sensor in sufficient concentrations. The range for vape detection and other sensors is also dependent on airflow, ventilation, filtration systems, and other variables. In most cases, alert thresholds for all sensors can be adjsuted to meet the unique needs of your environment.

Considerations

  • Verkada air quality sensors are not a life safety device, and should not be used as a smoke detector. Learn more about Detect and Understand Carbon Monoxide.

  • The SV21 lacks the sensors required to do vape detection and does not support this feature.

  • The SV11, SV23, and SV25 Vape Index measures air quality events indicative of vaping and smoking, but cannot provide 100% proof of a vaping incident.

    • School security and administrative teams should use the Vape Index and Verkada camera integration to help with investigations and monitor vaping activity and patterns, but use searches for physical evidence as the basis for further disciplinary/legal actions.

  • Certain smoke or cooking fumes can be identified as a vape event. In practice, this is rare in typical school environments.

  • The SV11, SV23, and SV25 will be less able to detect vaping/smoking incidents if a person uses significant efforts to hide the activity, such as exhaling into a closed container, out of a window, and so on. However, the more typical behavior of breathing into a shirt or jacket still results in a detected vape event in our testing.


Need more help? Contact Verkada Support.

Did this answer your question?