GEMINI-PERPLEXITY A.E.G.I.S.
(Atmospheric Early Grid Inspection System / Atmosferik Erken
SYSTEM REQUIREMENTS DOCUMENT (SRD)
Project Name: HIVE SCOUT v2.0
System Class: Industrial / Defense Grade Atmospheric Sensor Network
Document Status: Frozen Baseline Architecture
1.0 SCOPE AND OPERATIONAL OBJECTIVES
1.1 System Definition: HIVE SCOUT is an autonomous, fault-tolerant, networked atmospheric analysis system designed to detect biological/chemical aerosol risk clusters in urban canyons and stagnant air conditions. The system uses a tracer gas (Isobutylene) to detect physical air traps.
1.2 Out-of-Scope Functions: The system does not directly detect viruses, bacteria, or specific DNA/RNA. The system maps “risk zones” by measuring the suspension potential of particles using statistical and fluid dynamics vectors.
1.3 Operational Environmental Conditions: Field nodes will be designed to IP67 standards to operate outdoors in severe meteorological conditions, within a temperature range of -20°C to +50°C, and 0% to 100% relative humidity (including condensation).
2.0 HARDWARE AND SENSOR SPECIFICATIONS
2.1 Core Processing Unit: Each field node will utilize a microcontroller featuring an ARM Cortex-M4 (or equivalent) hardware FPU to perform edge computing and solve floating-point sensor equations in milliseconds.
2.2 Sensor Tolerance and Calibration Matrix:
Minimum performance criteria for the sensors are as follows:
| Sensor Type | Detected Target | Limit of Detection (LOD) | Margin of Error | Critical Environmental Blind Spot |
|---|---|---|---|---|
| PID (10.6 eV Lamp) | Isobutylene (C_4H_8) | 1 ppb | \pm 2% | \geq 95% Relative Humidity (Condensation) |
| Dual Channel NDIR | Carbon Dioxide (CO_2) | 400 ppm | \pm 15 ppm | Temperatures outside -20°C to +50°C |
| Capacitive Polymer | Relative Humidity (RH) | 0% | \pm 1.0% | Direct condensation on the sensor surface |
| Ultrasonic Anemometer | Wind Vector (\vec{v}) | 0.01 m/s | \pm 2% | Heavy rain (Signal refraction) |
| 2.3 Mechanical and Active Purging Design: Passive air intake will not be accepted. Each device will feature a micro-compressor / turbine fan system that draws air into the sensor chamber at a constant flow rate and forcefully evacuates it using back pressure in the event of saturation. |
3.0 MATHEMATICAL DEBUGGING (COMPENSATION) ALGORITHMS
Raw sensor data must be compensated at the hardware level using the following formulas before being transmitted to the center.
3.1 PID Sensor Humidity Quenching Function:
Correction function to be calculated instantaneously to prevent high humidity from quenching UV rays:
3.2 NDIR Ideal Gas Compensation:
Correction filter to be used to prevent deviations in optical density caused by temperature and pressure changes:
4.0 EMBEDDED SOFTWARE AND STATE MACHINE (FIRMWARE)
The system will never fall into an undefined state in the field. The operating logic of the device is deterministically limited to the following 6 states:
4.1 STATE_INIT (Initialization): Power-On Self Test (POST) is executed upon power-up. Memory and sensor buses are pinged.
4.2 STATE_WARMUP: The system waits until the optical and thermal sensors reach calibration temperature. No data is sent to the center.
4.3 STATE_NORMAL (Operational): Compensation algorithms run, data is collected, the risk score is calculated, and the payload is pushed to the center via telemetry.
4.4 STATE_PURGE (Active Purging): If dust, condensation, or exhaust gas saturation is detected in the sensor chamber, the data stream is cut, and the chamber is forcefully evacuated via back pressure.
4.5 STATE_DEGRADED (Limited Operation): If supporting sensors like humidity or temperature fail, they are bypassed via software. Risk calculation continues with a reduced reliability score.
4.6 STATE_FAULT (Critical Failure): In the event of a power collapse or main sensor failure, the system shuts down all hardware, sends an “SOS” packet to the center, and locks itself in Deep-Sleep mode awaiting a hardware reset.
5.0 SYSTEM SAFETY AND “RED LINE” RULES
5.1 Active Cross-Referencing Requirement: The system will not generate a Red Risk (Aerosol Clustering) alarm unless the following three conditions are verified within the exact same millisecond:
- Wind speed must be below the threshold (Stagnant air).
- There must be a positive increase gradient in the ambient carbon dioxide (CO_2) level.
- The tracer gas (Isobutylene) signal must remain stable without dispersing within a specific time window.
5.2 Watchdog Timer (WDT) Requirement: The microcontroller loop must feed the hardware Watchdog Timer upon every successful MQTT packet transmission or state transition. If the loop is broken or frozen, the device will automatically perform a hardware reset.
5.3 Fault Isolation: No processing is done with corrupted data. When faulty hardware is detected, the system automatically triggers a transition to STATE_DEGRADED or STATE_FAULT.
Bir yanıt yazın