Ask for both sides of accuracy
Detection rate measures how often a system flags approved vape tests. False-alarm rate measures how often it flags ordinary room events. A vendor can improve one by changing the threshold and make the other worse. Ask for both at the same setting.
Common room changes to test
- Deodorant and body spray
- Perfume and air freshener
- Cleaning spray and alcohol products
- Steam and sharp humidity changes
- Dust, construction work and HVAC changes
Do not assume every alert from these sources is a “false positive.” Some products label masking events apart from vape events. Confirm what staff will see on the phone and in reports.
How systems limit nuisance alerts
Makers may use several particle sizes, chemical sensors, room baselines, threshold tuning and software patterns. These can help. They do not remove the need for good placement and a local test.
A fair school test
Use the same room, distance, airflow state and event plan for each vendor. Run enough tests to see repeat results. Log missed events, wrong labels and nuisance alerts. Keep the staff receiving alerts blind to test time when possible. Save the settings used so a vendor cannot show one tuned demo and ship another setup.