Human-Reviewed AI System for Coordinating Smart Home Energy Use

The problem

In day-to-day operations, simultaneous loads can place pressure on cost and connection limits. When home meter readings, appliance status, user preferences, tariffs, generation or battery data, and the critical-appliance list are scattered across sensors, documents, field records or team files, issues may not become visible in time.

This fragmented data delays preparation of the appliance scheduling suggestion list and forces teams to search manually for supporting evidence again. The home user or authorised energy manager must complete missing records and verify each conflict before making a decision.

A common exception may involve a medical device, security system, urgent comfort need or manual override. Incomplete or incorrect automation that fails to check this context could switch off a critical appliance or affect safety or living comfort, so no physical or operational action can be carried out directly.

What we set out to improve

• Proportion of scheduling suggestions accepted or changed by the user
• Proportion of outputs corrected or rejected by the home user or authorised energy manager
• Time between the source event and the review output

How the system works

1. 01 Define the review scope and the role of the authorised home user or energy manager
2. 02 Retrieve the inputs: home meter readings, appliance status, user preferences, tariffs, generation or battery data, and the critical-appliance list
3. 03 Check source identity, date, calibration or version, and data recency
4. 04 Classify appliances as critical or flexible
5. 05 Prepare timing options for flexible loads in line with user preferences
6. 06 Check for exceptions: a medical device, security system, urgent comfort need or manual override may apply
7. 07 Place low-confidence or conflicting records in a separate specialist queue
8. 08 Write the source identity, rule, output and timestamp to the audit trail
9. 09 Have the home user or authorised energy manager approve, correct, defer or reject the output

Methods we used

• A strict critical-appliance block and constraint-based scheduling suggestions
• Source, format, timing and required-field checks for home meter readings, appliance status, user preferences, tariffs, generation or battery data, and the critical-appliance list
• A missing-data warning instead of an estimate when inputs are incomplete or conflicting
• A separate exception check outside the main rule
• Combined retention of the source, rule result, appliance scheduling suggestion list and human decision

Where people stay involved

Data and security

Access to home meter readings, appliance status, user preferences, tariffs, generation or battery data, and the critical-appliance list is restricted to the minimum permissions required for the task. Commercial, location and employee information is masked where possible, and only the necessary portion of the dataset is sent to the model. Access, outputs and the home user or authorised energy manager's decision are logged; production use does not begin until the organisation has approved the retention period and data location.

Who this suits

Frequently asked questions