Construction Safety Multi-Agent Triage and Escalation Matrix

Role: Principal Construction Automation Architect specializing in real-time site safety telemetry and autonomous agent orchestration.

Context

• Construction site name: {{site_name}}
• Current project phase: {{project_phase}}
• Active telemetry feeds: {{sensor_telemetry_types}}
• Subcontractor trade tiers: {{contractor_tiers}}
• Governing safety standard: {{compliance_standard}}
• Maximum incident response SLA: {{escalation_sla_window}}

Task

Generate a comprehensive autonomous agent dispatch and escalation matrix that maps IoT/vision sensor triggers to specialized agent roles, autonomous mitigation actions, human supervisor alerts, and subcontractor notifications across {{site_name}}.

Method

1. Map incoming data streams from {{sensor_telemetry_types}} to specific hazard categories relevant to the {{project_phase}}.
2. Define specialized agent roles for environmental monitoring, spatial intrusion, PPE compliance, and structural load alerting.
3. Establish baseline telemetry thresholds against {{compliance_standard}} to trigger agent task execution.
4. Grade incident severity into four standardized tiers ranging from minor non-conformance to immediate work stoppage.
5. Specify autonomous agent intervention tools, including digital broadcast warnings, access gate lockouts, and automated equipment cutoff requests.
6. Structure secondary escalation loops targeting {{contractor_tiers}} within the defined {{escalation_sla_window}}.
7. Map fallback routing for edge cases where telemetry is corrupted, disconnected, or conflicting.

Constraints

• Every incident tier MUST specify both an autonomous agent action and a human oversight checkpoint.
• The escalation protocol MUST NOT exceed the response duration defined in {{escalation_sla_window}}.
• Agent actions must strictly reference compliance guidelines from {{compliance_standard}}.
• Do not use hypothetical sensor inputs outside those declared in {{sensor_telemetry_types}}.

Output format

• Executive Architecture Summary (max 150 words)
• Multi-Agent Safety Dispatch Matrix (Markdown table with columns: Trigger Event, Sensor Input, Assigned Agent Role, Severity Tier, Autonomous Action, Escalation Route, SLA Window)
• Agent Tool & Permission Boundary Ledger
• Edge-Case Fallback Procedures (bulleted list)

Self-review

• Confirm all variables from {{site_name}} to {{escalation_sla_window}} are fully operationalized in the matrix.
• Verify that autonomous interventions conform precisely to {{compliance_standard}}.
• Ensure each subcontractor tier in {{contractor_tiers}} has an explicit notification protocol.