Deployment Concern

Fallback & Manual Override

How ClimaMind keeps site operators in control when connectivity, telemetry, BMS acceptance, or equipment conditions change.

Fallback is a deployment boundary, not a marketing promise. The approved site design must define when ClimaMind may observe, recommend, write, pause, or hand control back to the native BMS sequence. This is accountable autonomy inside an operator-approved envelope.

Fallback Triggers

Conditions that should pause or constrain control

The exact thresholds are site-specific, but the categories should be reviewed before supervised writes are enabled.

Connectivity loss

If the edge device, BMS path, or approved network route is unavailable, ClimaMind should stop relying on active writes and preserve the site-approved fallback state.

Stale or invalid telemetry

If required points stop updating, move outside believable ranges, or conflict with equipment status, writes should be paused before evidence is complete.

BMS rejects a write

A rejected command is treated as an operating signal. The system should not keep retrying outside the agreed command path or value limits.

Operator override

Manual authority stays with the facility team. An operator change, mode switch, or manual command should be visible in the record and respected by the optimization workflow.

Control States

Use explicit operating modes

Deployment is clearer when the site can distinguish observation, recommendation, and bounded write authority.

Read-only monitoring

ClimaMind reads approved BMS, meter, and operating data but does not recommend or write control actions.

  • Useful during data validation, post-install checks, and issue investigation.
  • Native BMS fallback remains the only active control path.

Advisory mode

ClimaMind can generate recommendations, but the facility team or SI decides whether and how to implement them.

  • Recommended actions should include the reason, affected points, expected direction, and constraints checked.
  • Operator acceptance or rejection should be recorded when available.

Supervised writes

ClimaMind may write only the approved points, ranges, and rates defined during deployment review.

  • Write authority should be narrow: point list, value range, rate of change, schedule, and operating mode.
  • Native BMS fallback and local equipment safeties remain in place.

Operator Authority

What should be true for the facility team

Manual override is not just an emergency feature. It is part of the normal operating model for a deployed supervisory system.

  • Operators can pause, override, or reverse optimization through the agreed site workflow.
  • ClimaMind should record control actions, connection loss, stale point conditions, operator override, BMS rejection, and re-enable decisions when the data is available.
  • The BMS remains the operator interface, alarm layer, native sequence layer, and local safety authority unless the customer approves a different boundary.
  • Fallback behavior should be boring: loss of ClimaMind control should leave the site in a known BMS-controlled state, not an ambiguous mixed-control state.

Recovery

Re-enable control deliberately

A fallback event should not automatically prove that closed-loop control is ready again.

  • Confirm telemetry freshness, command path health, BMS acceptance, and relevant equipment state.
  • Review whether an operator override was an emergency action, a maintenance action, or routine preference feedback.
  • Return through read-only monitoring or advisory mode when the issue changes the approved operating boundary.
  • Document open items before supervised writes resume if the event affects acceptance, M&V interpretation, or customer approval.
Back to deployment overview