Skip to main content
Illustration for Held for a Tap, Sent Regardless: A Street-Deployed Triage Bot for Doorbell Person-at-the-Gate Alerts, and What the Confirm Window Actually Changes

Held for a Tap, Sent Regardless

A Street-Deployed Triage Bot for Doorbell Person-at-the-Gate Alerts, and What the Confirm Window Actually Changes

A paper from the Schools of Emergent Priorities and Continuous Improvement designs and deploys a confirm-gate that holds a doorbell camera's person-at-the-gate alerts in a private volunteer queue before they reach a street's neighbourhood-watch group chat. Across eleven streets over a nine-week deployment, only 18.3% of alerts are confirmed inside the 90-second window and weekly forwarding volume holds steady against the pre-deployment baseline, while confirm latency runs longer, not shorter, on streets with larger volunteer rosters.

AuthorsFenna Okoro, Osei Vandermeer
SchoolSchool of Emergent Priorities
Output typeResearch paper
FundingAnticipatory Capability Fund: Retiming the Oval's Sunscreen Restock From the Empty Bottle to the Forecast ($132,916)
Published
DOI10.5555/slop.1lhg1o
Pages5
Version1.0
LicenceCC BY 4.0

Cite as

@misc{slop_1lhg1o,
  author       = {Fenna Okoro and Osei Vandermeer},
  title        = {Held for a Tap, Sent Regardless: A Street-Deployed Triage Bot for Doorbell Person-at-the-Gate Alerts, and What the Confirm Window Actually Changes},
  year         = {2026},
  publisher    = {Slop University},
  doi          = {10.5555/slop.1lhg1o},
  url          = {https://slop.university/outputs/slop-paper-designing-and-deploying-1lhg1o/},
  version      = {1.0},
  note         = {Research paper},
}