Duress Response

What you can build

• Confirm a silent duress press before anyone calls

  A cashier presses a wireless panic button and Verkada raises a silent alarm. The operator proposes a Slack notice with the nearest camera still and the last door badge, then the first call. A person approves each.

  With Verkada, Slack

• Run the call chain from a lone-worker button

  A LoRaWAN button in a plant room uplinks a press. The operator proposes calls through Amazon Connect in the order your plan lists, one at a time, and a page to the duty officer through PagerDuty.

  With LoRaWAN Network, Amazon Connect, PagerDuty

• Leave one incident record instead of five screenshots

  Alarm, still, footage link, badge event, calls placed, and every approval land in one record when Genetec Security Center raises a duress event. The record closes when a person confirms the scene is clear.

  With Genetec Security Center, Microsoft Teams

Requirements

• One access control or video connector that raises panic or duress alarms, such as Verkada, Genetec Security Center, or Milestone XProtect
• A camera and door map naming which camera and which door sit nearest each button
• One voice connector, Amazon Connect or Twilio, and one messaging connector: Slack, Microsoft Teams, PagerDuty, or xMatters
• A written call chain: who is called, in what order, and what the person answering should hear
• A LoRaWAN Network connector if your buttons are LoRaWAN devices rather than inputs on the alarm panel

Authentication

It acts through the API keys, webhook secrets, and calling credentials your security, sensor, voice, and messaging connectors already hold. It is issued none of its own.

Limits

• It confirms a press with evidence. It does not decide that an alarm is false, and it never cancels a monitoring dispatch.
• The still is the frame nearest the trigger time. Where retention or a camera outage leaves a gap, the record says so.
• Outbound calls are paced by your voice platform's own limits. A long chain takes longer than a short one.
• It works from your button-to-camera map. A button not on the map is proposed with the alarm alone, without a still or a door event.

Questions and answers

What does a person approve?

Each call, each notice, each page, and the incident record. The approver sees the alarm, the still, the door event, and the exact number or channel the step would use. Approve, edit, or dismiss each one. Nothing is dialed or posted until then.

What record is left after an alarm?

One incident record per alarm. It holds the alarm identifier, the device, the still, the footage link, the badge event, every call and notice sent, who approved each, and when the platform resolved the alarm. Each step keeps a receipt: what was sent, why, and how to undo it.

Does it ever lock a door, cancel an alarm, or call the police on its own?

No. It reads alarms and evidence. Calls and notices go out only after a person approves them. It never releases or locks a door, never resolves an alarm in the platform, and never contacts a monitoring center or the police. Those stay with your plan and your people.