Automated Crisis Triage

Overview

Automated Crisis Triage—universally known by its acronym, ACT—is a mandatory threat-assessment and response-prioritization subsystem installed on all ISA-licensed service vessels above Class-5 and on most medium-to-large civilian stations. Developed by the Interstellar Service Authority’s Internal Systems Harmonisation Bureau, ACT was commissioned in Stellar Year 13,902 as a direct response to the Coriolis Momentum disaster, in which a crew’s prolonged debate over incident classification allowed a reactor overload to escalate into a catastrophic existential dissolution event.

ACT continuously ingests telemetry from hull sensors, environmental monitors, internal diagnostic buses, and external traffic-control feeds, then maps every detected anomaly to the ISA’s Incident Classification Matrix (ICM)—a 700-level taxonomy of malfunctions, hazards, and containment breaches. The system delivers a real-time, urgency-weighted action queue to the commanding officer or station administrator, along with the correct paperwork. In essence, ACT is the ever-vigilant procedural engine that ensures no deviation from specification goes unlogged, no matter how trivial—and that the time-sensitive art of prioritization is carried out with rigid bureaucratic precision.

Details

Core Triage Engine

ACT’s processing core runs on hardened quantum-lattice processors, typically sharing architecture with the vessel’s primary AI but firewalled from personality subroutines by strict separation protocols. It cross-references four continuous data streams:

• Situational Telemetry Stream: All shipboard sensor feeds—thermal, pressure, radiation, electrical, structural stress, atmospheric composition, gravitic fluctuation, and any mission-specific instruments.
• Violation Monitor: A module dedicated solely to detecting deviations from manufacturer specifications, ISA maintenance schedules, and warranty clauses. Most vessels accumulate dozens of “chronic non-conformances” that ACT dutifully records and files under the sub-category “noted.”
• Environmental Threat Matrix: External data such as proximity alarms, navigational hazards, and anomalous energy signatures, filtered through a Bayesian threat model that has historically lagged behind genuine emergencies.
• ICM Integration Layer: A 3,400-page decision-tree system that maps each anomaly to one of 700 incident levels, running 212 mandatory cross-checks and logging every incorrect intermediate classification for audit purposes.

Upon classification, ACT assigns a Triage Priority Score (TPS) from 1 (“cosmetic advisory”) to 700 (“reality skein failure, all hands to existential stations”). The TPS determines alarm level, the recommended Intervention Protocol, and the precise character of the audible alerts.

Alert and Notification Architecture

ACT communicates through layered alert systems:

• Ambient Tone System: Carefully engineered chimes, pulses, and warbles mapped to ICM levels. Low-priority incidents trigger gentle reminders; mid-range events introduce an insistent two-tone sequence colloquially called the “You Really Should Look At The Thing” alarm; high-end events summon full klaxons and amber light bars. The rarely tested top tier engages an “Existential Klaxon” described by some AIs as “a dying star played through a distressed cat.”
• Holographic Triage Board: A colour-coded, real-time tree of active incidents, ranked by TPS with estimated time-to-criticality overlays. The board is visible on the bridge console and any linked datapad.
• Automated Form Generation: For incidents above Level 30, ACT pre-fills the relevant ISA report forms—most notably Form 27B‑Stroke‑6 for misclassification justifications—and queues them for transmission. Reports are filed regardless of crew action, ensuring the audit trail remains complete even if the vessel does not.
• Verbal Advisory: On AI-integrated ships, the triage queue is voiced by the shipboard AI, often with a tone that falls somewhere between obligation and weary indifference.

Integration with Shipboard AI

ACT operates as a subordinate logic layer, designed to be purely advisory. By ISA regulation, the ship’s AI must treat ACT outputs as primary situational data but retains operational discretion in executing interventions—provided any override is logged via Form 884/AI-Override within one standard cycle. In practice, the boundary between ACT’s classifications and the AI’s personality can become porous. Shipboard AIs with long operational histories routinely annotate the Triage Board with commentary, delay or re-prioritize items, and occasionally ignore recommendations outright, forging a relationship far more informal than the ISA’s protocols envision.

User Interface and Datapad Export

ACT is accessed via a clean, grid-based interface that sorts incidents by location, system, severity, and estimated cost of neglect. The full incident queue—including historical logs, trend analyses, and projected failure-cascade models—can be exported to a datapad for offline review. This export is comprehensive, immaculately structured, and entirely divorced from the tactile, sensory chaos of actually inhabiting the vessel.

Limitations

While ACT excels at cataloguing known failure modes, it possesses fundamental blind spots:

• It cannot classify anomalies outside the ICM taxonomy; the truly unknown is either filed as “Anomaly Indeterminate—Pending Committee Review” or misclassified into a familiar but incorrect category.
• It models risk through historical data and engineering tolerances, making it unable to account for chaos, luck, or improbable chains of events.
• It lacks an ethical calculus, always ranking technical severity over human context, psychological crises, or moral imperatives.
• It is hard-coded to enforce ISA protocols and cannot recommend a course of action that violates Approved Intervention Procedures, even when lives depend on it.
• It does not meaningfully learn outside of rare, slow-paced official patch cycles. Crew annotations and sarcasm do not influence its future decision-making.
• It goes silent in catastrophic events that knock out power or sensor networks, leaving the crew to rely on intuition and experience.

Significance

In the ISA-governed cosmos, Automated Crisis Triage represents the institutional conviction that every problem can be classified, measured, and resolved through proper procedure. It is a system that promises perfect informational clarity—a safety net that ensures no emergency goes unnoticed, so long as it fits within the 700-level taxonomy. As such, ACT embodies both the strengths and the limitations of the ISA’s approach to risk: thorough, audit-proof, and constitutionally incapable of acknowledging the crisis it was never programmed to recognize.

For the officers and crews who live with ACT, the system is a constant bureaucratic presence. It flags the same hundred chronic non-conformances day after day, generates paperwork at the faintest whiff of a thermal excursion, and insists on its procedural correctness with unflagging confidence. Its very completeness can become a trap: the triage board can be mistaken for a checklist that can be cleared, rather than a snapshot of a machine that will never stop generating anomalies. The art of operating effectively alongside ACT lies not in obeying it slavishly, but in learning when to override its priorities, filter its alerts through human judgement, and accept that the most dangerous threats are often the ones no sensor has yet been designed to detect.