Rimward
EN
EN · EnglishFR · Français
Request a site assessmentSite assessment
☰
Rimward · Escalation research · September 2026

What happens between an alert and intervention?

A nearby station is easy to measure. A usable escalation process is harder: someone must assess the evidence, choose an appropriate action and establish what happened next. Public geography and French verification rules help distinguish those responsibilities—but do not supply response times or cost savings.

  • 2,808 solar-register entries ≥ 1 MW
  • 2,319 comparable GN entries
  • French public sources
  • Snapshot: 30 September 2026
99.5%Selected entries assigned to Police or Gendarmerie

The geographic join does not reveal a fleet mostly outside public-law-enforcement responsibility.

Of 2,808 selected electrical entries, 2,321 are assigned to Gendarmerie Nationale (GN) and 473 to Police Nationale through their implantation communes. Fourteen assignments remain unresolved; that is not evidence of an absence of public responsibility.

An institution assignment is not a record of monitored cameras, staffed verification or guaranteed intervention. Those are separate operating questions. The useful distinction is not public versus private protection, but which work each arrangement actually provides.

Public responsibility is identifiable; service delivery is not

The official table links each commune to an institution and service. It does not describe a portfolio's monitoring contracts, an available patrol or an arrival commitment.

Official institution assignment · all 2,808 selected electrical entries

  • Gendarmerie Nationale2,321 82.7% of the selected entries
  • Police Nationale473 16.8% of the selected entries
  • Unresolved assignment14 0.5% · missing, blank or conflicting institutions
Assignment uses the implantation commune, not the delivery address. These are register entries, not validated physical sites. Police-assigned and unresolved entries are outside the following GN-only distance comparison.

For an owner, the practical task is to keep actual access information, the agreed escalation procedure and current contacts together. A responsibility lookup helps with coordination; it cannot replace the operating procedure.

The closest address is not a dispatch model

For 506 of 2,319 comparable GN entries—21.8%—even the closest published location of an assigned service is farther from the commune centre than the nearest location in the comparison pool. The pool contains directory locations whose service ID appears somewhere in the GN responsibility table.

Median straight-line commune-centre distance, km · same 2,319-entry GN subset

  • Nearest public-directory GN location4.7 km P90: 9.1 km · includes non-territorial directory entries
  • Nearest GN location in the comparison pool4.9 km P90: 9.6 km · service ID used in the responsibility table
  • Assigned services: closest published location5.4 km P90: 11.3 km · services assigned to the entry's commune
Distances start at commune centres, not plants or patrol positions. P90 marks the distance within which 90% of compared entries fall; it is not a confidence interval. The median difference is not a typical extra journey, and equal centre distances do not prove co-location.

The Ministry's description of the 17 emergency service makes the missing layer explicit: the call centre assesses the report and can send teams, redirect the call or give instructions. A station address does not encode that decision or resource availability.

Geography remains useful for coordination and site access. To evaluate intervention, an owner needs measured requests, acceptance, arrival or missing outcomes—not a kilometre-to-minutes conversion.

Verification is work; attendance is one possible action

For remote property surveillance, Article L613-6 addresses unjustified calls leading to undue public intervention when not preceded by checks of consistent indications of a flagrant crime or offence. That puts evidence assessment in the operating chain, without proving that every alert needs a private visit.

Decision framework · not measured timings, action rates or savings

Assess the evidence and apply the customer's procedure

Establish the event context, actual location and available evidence. Unusable or absent imagery leaves an unknown—not proof that an alert is false.

  • Remote action, where appropriate

    An approved audio warning or other configured action may be available at an equipped site. Its suitability and observed effect must be checked; it is not automatically a substitute for attendance.

    Record: attempt, observed reaction, follow-up and unknowns.

  • Contracted private attendance

    A private responder may be one route under the agreed procedure. Coverage windows, access, callout charges and arrival commitments need to be explicit.

    Record: request, acceptance, arrival, outcome and fees.

  • Public emergency escalation

    Immediate or serious facts follow the emergency procedure. The public service assesses the call; territorial assignment is not a contractual response-time guarantee.

    Record: information supplied, receipt and known subsequent actions.
Paths can run in parallel; this is not a mandatory serial sequence. Do not delay urgent assistance while attempting a remote challenge. We measured neither remote resolution nor private/public response outcomes.

Remote verification or a suitable remote action could avoid an unnecessary journey in some cases; neither the directory nor this analysis establishes how often. A person leaving a camera view after a warning is an observation, not by itself proof that a theft was prevented.

This distinction belongs in the budget. Staffed review, usable evidence, any remote-action capability and contingent private attendance are different cost components. Price their scope and availability rather than assuming either that every alert needs a visit or that no visit will be needed.

Ask for a measurable escalation service

The public study identifies what must be checked; operating records must establish what the service achieves.

  1. Can the operator verify the event under actual site conditions?

    Test usable day/night evidence, access failures and the agreed fallback. Record missed or failed tests and unknown coverage, not only successful alerts.

  2. Which actions are available, authorised and priced?

    Confirm operator responsibilities, configured remote actions, private callout terms and emergency criteria. Do not infer any of these from a nearby station.

  3. What evidence supports a claim of fewer journeys?

    Require a defined event population, observation period, attempted actions, outcomes, unresolved cases and actual attendance. Compare like-for-like procedures; a lower dispatch count alone may conceal unresolved events.

Fund the chain you can verify

These data do not establish whether public-service delivery is adequate, nor justify automatic private dispatch. Public responsibility, remote assessment, optional remote action and contracted attendance describe different roles.

For a portfolio decision, establish those roles and cost components, then test the agreed route and review its records. This analysis provides no measured deterrence rate, avoided-dispatch saving or guarantee of intervention—and does not establish such benefits for any vendor.

Sources, calculation and limits

Geographic snapshot
RTE edition 31 July 2026; register, GN directory, competence table and commune centres fetched 30 September. The same 2,808-entry selection as the fleet study uses implantation commune first; no delivery-commune fallback is needed here. Six entries lack a centre match. Of 2,321 GN-assigned entries, 2,319 have a complete distance comparison.
Directory and distance
Exclude ten rows with unusable coordinates. Keep 3,670 distinct service-ID/location pairs; 3,037 pairs belong to IDs used in the GN responsibility table. Use great-circle distance from the official API Géo centre and the closest location of any assigned service. A distance difference greater than one metre counts as farther; this does not measure separation between stations. P90 uses nearest rank.
Weighting and approximation
The 21.8% proportion becomes 23.3% with one observation per commune and stays 21.8% without Bas-Rhin or in metropolitan entries alone. Repeated commune locations are not independent site positions. Actual-plant proxy error, road routes and patrol positions were not validated. The remaining 78.2% is not proof of a correct nearest-service match.
Context is not outcome data
The French remote-property-surveillance rule and the Ministry's emergency-service explanation were checked on 30 September 2026. The decision framework is our synthesis, not an issued operational instruction or complete legal-compliance checklist. No private incident, audio-action, operator-time or dispatch-cost dataset was analysed.