UV theft detection powder application tips, paired with the camera zone that catches the touch.

Less than the product photos imply. A thin dusting that you can barely see in normal room light is the right thickness. As a working number, 0.2 to 0.5 grams per 100 cm² of treated area for a smooth interior surface (the inner shelf of a mailroom drawer, the underside of a cash tray) and 1 to 3 grams when mixed into petroleum jelly as a paste for outdoor or vertical surfaces (a transformer cage latch, a spool flange). A heaped pile is counterproductive. It looks suspicious in normal light, it transfers in clumps that wash off in the first hand-wash, and it produces a UV signature dramatic enough that the suspect will see it on their hand and know exactly when they touched the trap. Discreet quantity, applied with a soft brush or a feather, is the technique vendors describe and the technique field deployments confirm. Apply once, document the application log, and re-treat on the schedule below.

Worth treating: surfaces where legitimate contact is narrow and easy to enumerate (a mailroom inner shelf, a specific tool crib, the underside of a cash drawer divider, a transformer cage latch, a fire alarm pull, a designated keyboard, the rim of a copper spool flange). Wasted: anything with broad legitimate contact (general door knobs, hand rails, parking lot bollards, retail sales floors, common-area handles). The product page on GLO Effex and the Sirchie technical sheet both make the same point in passing: the powder fluoresces but the chemistry alone cannot tell a court when the contact happened. If your treated surface gets touched 200 times a day by people who have a right to be there, glowing hands prove nothing. Pick surfaces where the legitimate-contact list is shorter than the suspect list, and the powder earns its place. Anywhere else, the powder is a prop.

Indoor, low-touch surfaces hold a usable powder layer for 14 to 30 days under normal HVAC airflow. The thin dusting lifts gradually with passive air movement, even before anyone touches the surface. Outdoor surfaces vary by weather. A petroleum-jelly paste survives one or two light rain events but should be re-applied weekly during wet seasons. Cash drawers and till trays should be re-applied at every shift change, both for chain-of-custody clarity and because the per-surface contact frequency is high. The application log should record the re-application date for every surface. A common mistake is applying once, never re-applying, and assuming the trap is still active months later; in fact the powder is gone and the property is running cameras over a clean shelf.

The dwell threshold is how many seconds the camera zone has to see a person inside the polygon around the treated surface before it fires the alert. The right number depends on how long a thief actually stops to touch that surface, not on motion-sensor defaults. Mailroom shelf: 10 to 20 seconds (residents walking past take 3 seconds; a thief stops, scans, lifts). Copper spool rack on a jobsite: 30 seconds (filters out workers walking past). Cash drawer at till: 8 to 12 seconds (the till is touched briefly by employees, longer by someone substituting bills). Transformer pad access panel: 60 seconds (zero legitimate touch outside scheduled maintenance). Tool crib gang box: 20 to 40 seconds (depends on whether the crib has legitimate after-hours access). Set the threshold tight enough to fire on actual theft, loose enough to let normal traffic pass without firing.

Because the powder washes off. The single most common failure mode in field deployments is a recurring thief who hits the mailroom on Sunday night, gets a thin transfer of powder on their hand, washes it off in the apartment sink at 23:14, and is back at work Monday morning glow-free. The trap fired correctly. The powder transferred. The hand was inspected at 09:00 Monday and showed nothing. The case never starts. The application protocol is not just about the powder, it is about the timing chain that pairs the application with an alert that lands fast enough for an inspection inside the wash-off window. That is why every per-surface entry below has both a powder column and a zone column.

No. The capture point for the camera-zone half of this protocol is the DVR or NVR's HDMI multiview output, the same signal that drives the existing guard monitor. A Cyrano edge unit plugs into that HDMI, runs person detection across every tile, applies the zone polygon and dwell threshold per tile, and emits the alert to a WhatsApp ops thread in under 60 seconds. No camera replacement, no firmware involvement, no per-camera credentials. Works on any DVR or NVR with HDMI out: Hikvision, Dahua, Lorex, Amcrest, Reolink, Uniview, Swann, Night Owl, and their rebrands. Up to 25 tiles per unit. The cameras you have stay where they are. The application protocol is the operational layer; the edge unit is the alert layer.

Date, time, applicator initials, exact surface treated (down to the specific shelf, drawer, or panel), powder product and lot or batch number, application thickness or quantity, applicator type (brush, feather, nitrile-glove dab), the planned re-application date, and the camera zone identifier this application is paired with. Two reasons. First, chain of custody: a contemporaneous application log is what defeats the defense argument that the powder was on the surface when the suspect touched it for legitimate reasons. Second, the log forces discipline. Properties that skip the log apply once, forget which surfaces they treated, never re-apply, and have no idea three months later whether their traps are armed or not. The log is a one-page sheet, signed and dated. Either you keep it or you do not have an application protocol.

A 365 nm UV flashlight is the best general choice. 365 nm produces less ambient fluorescence from common substances (laundry detergent residue, toothpaste, tonic water, petroleum jelly) than the cheaper 395 nm flashlights, so the contrast between the powder transfer and ambient glow is higher and the inspection is less ambiguous. Brands commonly used in field inspection: Convoy S2+ in 365 nm, Vansky 100 LED 395 nm (only acceptable budget option), and dedicated forensic torches from Sirchie. Document the flashlight model and wavelength on every inspection report. A 395 nm flashlight is acceptable for confirming a positive transfer, but if a defense lawyer challenges the inspection, the prosecutor will want the inspecting officer to have used 365 nm. Inspect in low light, photograph hands and any held items, log the timestamp.

Faster than the suspect's next hand-wash. Field reports from mailroom and tool-crib deployments suggest 30 to 90 minutes is the realistic upper bound; past two hours, expected wash behavior has degraded the powder transfer below detection threshold. The protocol that works in practice: the camera zone fires, the alert lands on WhatsApp under 60 seconds later, the on-site attendant or contracted patrol is dispatched to intercept or to inspect the access points the suspect would have used to leave, and the inspection happens at the front desk or in the patrol vehicle within a single-digit number of minutes. If the suspect has already left the property, the inspection happens at the next encounter (the next shift, the next maintenance ticket, the next package pickup) and the chance of recovery is lower. Speed of inspection is the lever the property controls. Powder formulation is fixed; application timing is fixed; the camera-to-inspection latency is the variable that decides whether the case closes.