Solution · Threat detection
Alerts that get acted on, not muted.
Weapons, intrusion, after-hours behavior, vehicle anomaly, and gunshot detection, routed to the right human in seconds. The hard part is false-positive tuning, and that's the part Tec-Tel owns.
- NDAA-compliant
- Platform-agnostic
- 1,000+ deployments over 15 years
How can we help?
Tell us what you're working through. We'll route it to the right person.
Threat detection and proactive alerts are real-time AI-driven detection of weapons, intrusion, vehicle anomalies, and gunshots, routed to the right human inside seconds. Tec-Tel composes weapons and gunshot detection with VMS-native and camera-agnostic analytics. The hard part is false-positive tuning: systems that cry wolf get muted, so Tec-Tel scopes a tuning cycle into every install.
§01 What threat detection covers
Seven detection types. Pick the three or four that fit.
Threat detection turns camera and sensor feeds into structured, real-time alerts. Seven detection types cover most commercial and education buyers. Not every site needs every one; the consultation picks the three or four that match your incident history.
§02 Routing
How alerts route to the right humans.
A detection that fires into a void is wasted. Alert routing is the part most vendors gloss over and most installs get wrong. The rule of thumb: every alert type has a primary recipient, a fast fallback when the primary does not acknowledge, and an escalation path that reaches dispatch.
Three recipients dominate. Security operations take the alerts that need a security response: weapons, intrusion, forced entry. Facilities take the operational alerts: a propped door, a forklift in a pedestrian aisle, an after-hours delivery running early. Police dispatch is the last hop for confirmed threats, routed through a UL-listed central station or a verified-alarm partner.
Tec-Tel scopes the routing matrix at install: who gets weapons alerts, who gets after-hours intrusion alerts, who gets PPE alerts, and what the fallback path looks like when the primary recipient does not acknowledge. We test the matrix as a tabletop exercise before go-live.
§03 The economics
High-false-positive systems get muted, fast.
Every threat-detection install lives or dies on alert fatigue. Once the false-alarm volume climbs past what an operator can triage, the queue gets muted, the integration gets disabled, and the genuine alarm gets missed with the noise. A system that cries wolf is worse than no system, because the customer is paying for theater.
Traditional motion alarms make this worse: they fire on any pixel change above a threshold, so a leaf, a shadow, or a passing truck all trip the alarm. AI threat detection classifies the contents of the frame (a person, a vehicle, a weapon, a fall) and triggers on the classification, not the motion, which collapses the alert volume.
The fix is the tuning window: confidence thresholds, motion masks, schedule-aware suppression, per-camera enable/disable. Tec-Tel scopes a tuning cycle into every install, reviews alert volume on a regular cadence, and drives the false-alarm rate down before signing a system production-ready. The other half is right-sizing: not every camera needs every detection.
§04 Integration
Wired into mass notification and lockdown.
Modern threat-detection platforms expose webhooks or event APIs that other systems subscribe to. The integration sequence on a confirmed threat: detection fires, mass notification goes out, IP speakers and digital signage activate, access-control doors lock down, on-site response begins, the dispatch path completes.
The integration is technical and the testing is operational. Tec-Tel runs a tabletop drill with the customer's security team, facilities team, and (for K-12) school administration before the system goes live. The drill catches the routing gaps a configuration review misses. NSGP and FEMA grant scoring explicitly reward documented response plans of this shape.
§05 Cost framing
What the alert pipeline costs.
Two pricing shapes dominate. Per-camera, per-month licensing fits VMS-native and overlay analytics, where the cost scales with camera count and the detection tier you license. Per-site annual subscription fits high-end systems like weapons detection and entrance screening, where the price scales with throughput and detection complexity, not camera count.
The underlying camera and recorder hardware is a separate spend that scales with resolution, environment, and install complexity. Tec-Tel itemizes the analytics and the hardware as separate lines on every proposal, so the alert-pipeline cost is never buried inside a hardware bundle.
Questions buyers ask us
FAQ
- What's the difference between AI threat detection and traditional motion alarms?
- Traditional motion alarms fire on any pixel change above a threshold: a leaf, a shadow, a passing truck. The false-positive rate is so high that operators tune sensitivity down until the alarm becomes useless. AI threat detection classifies what's in the frame: a person, a vehicle, a weapon, a fall, a fight. The trigger is the classification, not the motion. Tuned correctly, false positives drop substantially compared to motion-only systems.
- How do alerts route to the right human?
- By zone, by shift, by detection type. A perimeter-intrusion alert at a distribution center goes to the night-shift operator first, then to a contracted guard if the primary does not acknowledge. A weapons detection at a school goes to the resource officer plus the principal plus a confirmed 911 dispatch. Routing rules are configured in the VMS or the alert-broker layer and tested with a tabletop exercise before go-live.
- What false-positive rate is realistic in production?
- Vendor specs typically claim under 5% for object detection and under 10% for behavior analytics in tuned environments. Real-world numbers are higher at first and drop as zones, masks, and confidence thresholds get tuned. The single biggest predictor of long-term success is the tuning window. Tec-Tel scopes a tuning cycle into every threat-detection install rather than declaring it done at go-live.
- Why do high-false-positive systems get muted so fast?
- Alert fatigue is real and well-documented. When the false-alarm queue climbs past what an operator can triage, people disable or mute the alerts. Once the queue is muted, the genuine alarm gets missed too. The economics of threat detection live or die on this metric. A system that cries wolf is worse than no system, because the customer is paying for it and trusting it.
- How does this integrate with mass notification and lockdown?
- Most modern threat-detection platforms expose a webhook or event API that fires a mass-notification system and triggers an access-control lockdown inside the same alert. The integration sequence: detection fires, notification goes out to mobile devices and IP speakers, doors lock, on-site response starts, dispatch path completes. Tested as a tabletop drill before any deployment.
- Can existing cameras carry the analytics or do we need new ones?
- Often yes. Camera-agnostic overlay platforms run on the existing camera fleet down to about 1080p / 15 fps, and most weapons-detection platforms support modern IP cameras. Vendor-tied analytics require their own cameras. Tec-Tel audits the existing fleet and only recommends replacement where the resolution or framerate genuinely can't carry the detection you need.
- What does threat detection cost on top of the camera install?
- Two pricing shapes. Per-camera, per-month for VMS-native analytics, scaling with the camera count and the detection tier you license. Per-site annual subscription for high-end weapons-detection and entrance-screening systems, scaling with throughput. The underlying camera hardware is a separate line, scoped to the resolution and environment each position needs. Tec-Tel itemizes the analytics and the hardware separately on every proposal.
- Does the system satisfy school threat-detection grant rules (NSGP, SVPP)?
- It depends on the program and the year. NSGP scoring rewards documented monitoring with a measurable response window and a documented response plan. SVPP funds capital costs including AI threat detection at K-12 sites in many states. We design the install against the specific year's program rules.
Book a walkthrough
Want a read on your alert pipeline?
The free consultation pulls the recent alert history off your existing system, sorts the false positives from the genuine triggers, and shows where the routing matrix is dropping events. We'll tell you which detections are worth turning on and which will turn into noise.
- Tell us how many sites you run and what's already in place. We'll show you what a build or upgrade looks like.
- Straight answers from the team that does the work. We're platform-agnostic, so you get the system that fits your sites, not one brand's catalog.
Since 2010 · 1,000+ deployments nationwide · ISN-accredited
How can we help?
What you're looking for, plus any details. We review it and follow up, usually the same day.