Domain Frameworks

Chapter 8 — Domain Frameworks

This chapter covers the two major diagnostic frameworks that encode Dibyendu De’s field experience into structured, computable form: the FRETTLSM 88-factor causal taxonomy and the 300-rule Imperfection Rule Library.

FRETTLSM — The 88-Factor Causal Taxonomy

Why a Taxonomy?

When a machine fails, the question “why?” demands a systematic answer. A reliability engineer confronting a complex failure must consider dozens of potential causes spanning mechanical, thermal, electrical, chemical, tribological, and human factors. Without a systematic framework, investigation defaults to the most obvious symptom — “the bearing failed” — while the actual root cause (contaminated lubricant, excessive thermal cycling, installation error) goes unidentified.

FRETTLSM provides that framework. It is the “periodic table” of failure causes. Just as the periodic table organizes chemistry by making relationships between elements visible, FRETTLSM organizes failure engineering by making relationships between causal factors visible and tractable. A reliability engineer can systematically walk through all 88 factors, asking: Is this factor active? Is it initiating, accelerating, or retarding? What evidence supports or refutes its involvement?

The Eight Categories

F — Force / Flow / Foundation (12 factors)

Forces are the most direct cause of mechanical failure. This category covers anything that loads, moves, or supports the machine.

ID	Factor	Default I/A/R	Activation Wt	Threshold
F001	Dynamic imbalance	I	0.85	>2x baseline at 1x
F002	Misalignment force	I	0.90	2x > 50% of 1x
F003	Pipe strain	I	0.70	Shaft deflection >0.025mm
F005	Foundation resonance	A	0.80	f_nat within +/-20% of f_run
F009	Axial thrust	I	0.75	Displacement >50% clearance
F011	Impeller hydraulic imbalance	I	0.65	VPF >3x baseline
F012	Rotor bow	I	0.80	Runout >0.025mm

Module detection: Module B rules AFB06 (unbalance), COUP01-09 (coupling), FND01-10 (foundation), FL01-10 (fluid/flow).

R — Reactive / Electromagnetic (11 factors)

Electrical and electrochemical phenomena that degrade machinery from the inside out.

ID	Factor	Default I/A/R	Activation Wt	Threshold
R001	Chemical corrosion	I	0.70	Corrosion rate >0.1mm/yr
R004	Hydrogen embrittlement	I	0.80	Hardness change >5 HRC
R005	Stress corrosion cracking	I	0.90	Any crack indication
R010	Chloride pitting	I	0.75	Cl- >200 ppm

Module detection: Module B AC motor rules (AC01-AC09) and DC motor rules (DC01-DC07).

E — Environment (11 factors)

External conditions that the machine cannot control but must endure.

ID	Factor	Default I/A/R	Activation Wt	Threshold
E001	High ambient temperature	A	0.60	>40C
E003	Dust/particulate ingress	A	0.65	Si >25 ppm in oil
E005	Adjacent equipment vibration	A	0.45	Foundation vib >1.5mm/s
E010	Seismic/blast vibration	I	0.60	>0.05g sustained
E011	Lightning/power surge	I	0.70	Voltage spike >1.5x nominal

Module detection: Module A process classification and environmental correlation.

T_TIME — Time / Aging / Fatigue (11 factors)

The cumulative burden of operation. Time does not cause failure directly, but it accumulates the damage from everything else.

ID	Factor	Default I/A/R	Activation Wt	Threshold
T001	Cumulative fatigue cycles	I	0.75	>80% of design fatigue life
T003	Start-stop cycles	A	0.70	>design starts (typ. 5000)
T008	Insulation aging	I	0.80	IR <5 MO or PI <1.5

Module detection: Module F reliability layer.

T_TEMP — Temperature / Entropy (11 factors)

Heat is both symptom and cause.

ID	Factor	Default I/A/R	Activation Wt	Threshold
TT002	Hot spot formation	I	0.85	Delta-T >15C from adjacent area
TT004	Differential thermal expansion	I	0.75	Clearance change >20%
TT009	Thermal shock	I	0.85	>5C/min
TT011	Entropy increase (SEDL)	A	0.70	SE >0.7

Module detection: Module A temperature rules (TM001-TM008). TT011 bridges directly to Module B.3 SEDL entropy analysis.

L — Lubrication / Wear (11 factors)

The lubricant film is the last line of defense between metal surfaces.

ID	Factor	Default I/A/R	Activation Wt	Threshold
L002	Grease depletion	I	0.85	Temp rise >5C + overdue
L003	Water contamination	A	0.80	>200 ppm
L005	Wrong lubricant	I	0.90	Viscosity >+/-15% of spec
L007	Under-lubrication	I	0.90	Temp rise + gSE increase
L011	Film breakdown	I	0.85	Lambda < 1.0

Module detection: Module B bearing lubrication rules, lambda protocol.

S — Surface Topology (11 factors)

The condition of contact surfaces where failure physically manifests.

ID	Factor	Default I/A/R	Activation Wt	Threshold
S002	Pitting	I	0.80	BPFO/BPFI in envelope
S003	Spalling	I	0.90	Broadband +6dB
S005	Brinelling	I	0.85	Bearing defect freqs
S008	Cavitation erosion	I	0.80	HF increase + low NPSH

Module detection: Module A feature extraction (kurtosis, crest factor, impulse metrics).

M — Material / Man / Method (10 factors)

The human element and material limitations.

ID	Factor	Default I/A/R	Activation Wt	Threshold
M004	Installation error	I	0.80	Vibration >1.5x pre-shutdown
M005	Maintenance error	I	0.75	Performance degradation
M006	Operational error	I	0.70	Alarms from operator action
M010	Foreign object damage	I	0.90	Sudden vibration increase

Module detection: Module B.2 trend reversal detection (post-maintenance regression).

Asset Class Templates

The 88 factors are not all relevant to every machine type. FRETTLSM provides asset class templates that activate the relevant subset with machine-specific weights.

Centrifugal Pump Template (41 of 88 factors):

Top 5 initiators: F002 (misalignment), S008 (cavitation), L007 (under-lubrication), S002 (pitting), L002 (grease depletion)

Top 5 accelerators: L003 (water in oil), R011 (erosion-corrosion), TT001 (thermal cycling), L004 (particle contamination), E003 (dust ingress)

Critical retarders to monitor: proper alignment (counters F002), adequate NPSH margin (counters S008), correct lubrication schedule (counters L002/L007), seal flush plan (counters R001/L003), vibration monitoring program (detects S002/S003 early), minimum flow protection (counters M006 deadheading).

Templates are also defined for gearbox (40 factors), electric motor (38 factors), fan (35 factors), and compressor (42 factors).

IAR Dynamics

Every FRETTLSM factor carries an IAR classification that can change based on context. A factor may be an Initiator in one failure scenario and an Accelerator in another. For example:

F002 (Misalignment) is an Initiator when coupling misalignment is the root cause of bearing overload
F002 (Misalignment) is an Accelerator when thermal growth shifts alignment on a machine that was correctly aligned at ambient temperature

The IAR classification determines the response strategy: eliminate Initiators (strategic), suppress Accelerators (tactical), strengthen Retarders (defensive). The RCM framework uses IAR classification to select the appropriate maintenance tier.

The 300 Imperfection Rules

What Makes Imperfection Rules Different

Unlike fault detection rules (Module B, which identify symptoms), imperfection rules identify structural weaknesses — the design, installation, operational, or process-interaction conditions that make failure inevitable regardless of how well the machine is maintained.

The key differentiator: instead of saying “bearing vibration is high,” RAPID AI can say “bearing vibration is high due to shaft overhang ratio violation and coupling misalignment design.” This transforms alerts into engineering intelligence.

Rule Structure

Each imperfection rule has a consistent structure:

Field	Description
rule_id	Unique identifier (IMP_RULE_001 through IMP_RULE_300)
equipment_type	Target machine (pump, motor, gearbox, fan, compressor, steam turbine, conveyor, hydraulic system)
imperfection_category	Design, Installation, Operation, or Process Interaction
evaluation_logic	Parameterized condition expression
severity_weight	Importance on a 1-10 scale
engineering_reason	Why this condition constitutes an imperfection
recommended_action	Specific corrective guidance

Four Rule Categories

Design Rules evaluate whether the machine’s geometry, loading, and material selection are adequate for the service:

Rule	Equipment	Logic	Severity	Reason
IMP_001	Pump	`overhang_length / shaft_diameter > 1.5`	8	Excessive overhang increases bending stress and bearing radial load
IMP_004	Pump	`npsh_available < npsh_required`	9	Insufficient suction head causes cavitation and impeller damage
IMP_005	Gearbox	`gear_mesh_peak > threshold`	6	Excessive mesh excitation indicates tooth profile error

Installation Rules evaluate whether the machine was assembled correctly:

Rule	Equipment	Logic	Severity	Reason
IMP_002	Motor/Coupling	`alignment_reading > tolerance`	7	Coupling misalignment transfers axial and radial forces
IMP_003	Pump	`piping_load > allowable_limit`	7	Pipe strain distorts casing, affecting internal clearances
IMP_013	Motor	`axial_vibration / radial_vibration > 0.6`	7	Ratio indicates coupling misalignment

Operation Rules evaluate whether the machine is being operated within its design envelope:

Rule	Equipment	Logic	Severity	Reason
IMP_010	Motor	`bearing_temperature > 85`	7	Operating above thermal design limit accelerates lubricant degradation
IMP_020	Compressor	`pressure_variation > threshold`	8	Pressure oscillation indicates surge proximity
IMP_025	Conveyor	`belt_speed_difference > 5%`	5	Belt slip wastes energy and indicates tension issues

Process Interaction Rules evaluate machine-to-machine effects:

Rule	Equipment	Logic	Severity	Reason
IMP_030	Hydraulic	`oil_particle_count > contamination_limit`	7	Contaminated fluid accelerates valve and actuator wear
IMP_035	Fan	`foundation_stiffness < required_stiffness`	7	Insufficient stiffness amplifies rotor dynamic forces

Execution Flow

Retrieve equipment metadata (type, design parameters, operating conditions)
Fetch sensor and operating data from the pipeline
Load the relevant rule set for the equipment type
Evaluate each rule’s evaluation_logic against the data
Flag imperfections where the condition is satisfied
Assign severity and confidence scores
Rank active imperfections by risk index (severity x confidence)
Store results and generate the imperfection analysis report

Coverage

The 300 rules span 8 equipment types:

Equipment	Approx. Rules	Key Focus
Pump	~65	Shaft overhang, NPSH, hydraulic balance, seal design
Motor	~55	Thermal design, insulation, bearing clearance, cooling
Gearbox	~45	Tooth geometry, lubrication design, alignment, backlash
Fan	~35	Foundation, blade geometry, ductwork interaction
Compressor	~30	Surge margin, valve timing, intercooler design
Steam Turbine	~25	Blade clearance, thermal growth, governor stability
Conveyor	~25	Belt tension, roller alignment, structural fatigue
Hydraulic System	~20	Fluid contamination, valve tolerance, seal integrity

Connection to FRETTLSM

Every imperfection rule maps to one or more FRETTLSM factors. The shaft overhang rule (IMP_001) maps to F003 (pipe strain) and F009 (axial thrust). The cavitation rule (IMP_004) maps to S008 (cavitation erosion). The bearing temperature rule (IMP_010) maps to TT002 (hot spots) and L011 (film breakdown).

This connection ensures that the imperfection analysis is grounded in the same causal taxonomy as the fault detection rules. When an imperfection is identified, the FRETTLSM classification tells the engineer whether it is an Initiator (eliminate it), an Accelerator (suppress it), or a missing Retarder (strengthen it). The IAR classification from Chapter 2 transforms the imperfection finding into a strategic maintenance decision.

Complete FRETTLSM Factor Catalog

The FRETTLSM framework provides a systematic diagnostic lens for root cause analysis. Each letter represents a category of failure mechanism that must be considered when diagnosing any fault. The framework prevents premature closure — the tendency to stop at the first plausible cause. When a bearing shows elevated kurtosis, FRETTLSM forces the analyst (and the engine) to ask: is this Force (overload)? Lubrication (starvation)? Surface (contamination)? Environment (corrosion)? Material/Man (wrong bearing installed)?

The Eight Categories with Factor Counts

Code	Category	Description	Factor Count	Universal?
F	Force / Flow / Foundation	Mechanical forces: imbalance, misalignment, looseness, hydraulic forces	12	Yes — all rotating machinery
R	Reactive / Electromagnetic	Electrical faults: eccentricity, bearing currents, winding degradation	11	No — motors only
E	Environment	External contamination, ambient conditions, corrosion	11	Yes — all equipment
T_TIME	Time / Aging / Fatigue	Cumulative operating burden: fatigue cycles, start-stop, insulation aging	11	Yes — all equipment
T_TEMP	Temperature / Entropy	Thermal effects: hot spots, differential expansion, thermal shock	11	Yes — all equipment
L	Lubrication / Wear	Oil/grease condition: depletion, contamination, wrong viscosity, film breakdown	11	Lubricated components only
S	Surface Topology	Contact surface condition: pitting, spalling, brinelling, cavitation erosion	11	Yes — all equipment
M	Material / Man / Method	Human factors, material selection, installation and maintenance errors	10	Yes — all equipment
	Total		88

F — Force / Flow / Foundation (12 Factors)

Forces are the most direct cause of mechanical failure. This category covers anything that loads, moves, or supports the machine.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
F001	Dynamic imbalance	I	0.85	>2x baseline at 1x	1x vibration, horizontal dominant
F002	Misalignment force	I	0.90	2x > 50% of 1x	Axial vibration, 2x harmonic
F003	Pipe strain	I	0.70	Shaft deflection >0.025mm	Directional ratio anomaly
F005	Foundation resonance	A	0.80	f_nat within +/-20% of f_run	Amplified vibration at specific speeds
F009	Axial thrust	I	0.75	Displacement >50% clearance	Axial displacement trend
F011	Impeller hydraulic imbalance	I	0.65	VPF >3x baseline	Vane pass frequency amplitude
F012	Rotor bow	I	0.80	Runout >0.025mm	1x vibration, phase-locked

Module detection: Module B rules AFB06 (unbalance), COUP01-09 (coupling), FND01-10 (foundation), FL01-10 (fluid/flow).

R — Reactive / Electromagnetic (11 Factors)

Electrical and electrochemical phenomena that degrade machinery from the inside out.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
R001	Chemical corrosion	I	0.70	Corrosion rate >0.1mm/yr	Oil analysis, visual inspection
R004	Hydrogen embrittlement	I	0.80	Hardness change >5 HRC	Material testing
R005	Stress corrosion cracking	I	0.90	Any crack indication	NDT inspection
R010	Chloride pitting	I	0.75	Cl- >200 ppm	Water/process chemistry

Module detection: Module B AC motor rules (AC01-AC09) and DC motor rules (DC01-DC07).

E — Environment (11 Factors)

External conditions that the machine cannot control but must endure.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
E001	High ambient temperature	A	0.60	>40C	Process monitoring
E003	Dust/particulate ingress	A	0.65	Si >25 ppm in oil	Oil analysis
E005	Adjacent equipment vibration	A	0.45	Foundation vib >1.5mm/s	Cross-machine correlation
E010	Seismic/blast vibration	I	0.60	>0.05g sustained	External event correlation
E011	Lightning/power surge	I	0.70	Voltage spike >1.5x nominal	Electrical monitoring

Module detection: Module A process classification and environmental correlation.

T_TIME — Time / Aging / Fatigue (11 Factors)

The cumulative burden of operation. Time does not cause failure directly, but it accumulates the damage from everything else.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
T001	Cumulative fatigue cycles	I	0.75	>80% of design fatigue life	Operating hours tracking
T003	Start-stop cycles	A	0.70	>design starts (typ. 5000)	Start counter
T008	Insulation aging	I	0.80	IR <5 MO or PI <1.5	Insulation resistance testing

Module detection: Module D reliability layer (Weibull parameter adjustment).

T_TEMP — Temperature / Entropy (11 Factors)

Heat is both symptom and cause. Thermal effects bridge directly to the SEDL entropy analysis in Module B.3.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
TT002	Hot spot formation	I	0.85	Delta-T >15C from adjacent area	Thermal imaging, bearing temp
TT004	Differential thermal expansion	I	0.75	Clearance change >20%	Alignment change with temperature
TT009	Thermal shock	I	0.85	>5C/min rate of change	Temperature rate monitoring
TT011	Entropy increase (SEDL)	A	0.70	SE >0.7	Spectral entropy computation

Module detection: Module A temperature rules (TM001-TM008). TT011 bridges directly to Module B.3 SEDL entropy analysis.

L — Lubrication / Wear (11 Factors)

The lubricant film is the last line of defense between metal surfaces. Lubrication failures are disproportionately severe — the highest severity scores in the rule set belong to lubrication starvation rules.

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
L002	Grease depletion	I	0.85	Temp rise >5C + overdue	Temperature trend + schedule
L003	Water contamination	A	0.80	>200 ppm	Oil analysis
L005	Wrong lubricant	I	0.90	Viscosity >+/-15% of spec	Oil viscosity testing
L007	Under-lubrication	I	0.90	Temp rise + gSE increase	HF amplitude + temperature
L011	Film breakdown	I	0.85	Lambda < 1.0	Lambda ratio calculation

Module detection: Module B bearing lubrication rules, lambda protocol.

S — Surface Topology (11 Factors)

The condition of contact surfaces where failure physically manifests. Surface damage is the second-most-common fault mechanism across the entire rule set (30% of all SENSE.CF rules).

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
S002	Pitting	I	0.80	BPFO/BPFI in envelope	Envelope spectrum analysis
S003	Spalling	I	0.90	Broadband +6dB	Broadband vibration increase
S005	Brinelling	I	0.85	Bearing defect freqs	Defect frequency matching
S008	Cavitation erosion	I	0.80	HF increase + low NPSH	High-frequency acoustic + process

Module detection: Module A feature extraction (kurtosis, crest factor, impulse metrics).

M — Material / Man / Method (10 Factors)

The human element and material limitations. These factors are detected through trend reversal analysis — watching for post-maintenance regression (the Waddington Effect).

ID	Factor	Default I/A/R	Activation Wt	Threshold	Observable Proxy
M004	Installation error	I	0.80	Vibration >1.5x pre-shutdown	Post-maintenance trend comparison
M005	Maintenance error	I	0.75	Performance degradation	Post-maintenance regression
M006	Operational error	I	0.70	Alarms from operator action	Event correlation
M010	Foreign object damage	I	0.90	Sudden vibration increase	Step-change detection

Module detection: Module B.2 trend reversal detection (post-maintenance regression).

FRETTLSM Distribution Across Component Fault Rules (119 Rules)

The distribution of FRETTLSM categories across the 119 SENSE.CF component fault rules reveals what types of energy flow disruption dominate in industrial rotating machinery:

Category	Rule Count	% of Rules	Insight
F (Force/Flow/Foundation)	51	44%	Mechanical forces dominate — most failures have a force imbalance root cause
S (Surface)	35	30%	Surface degradation is the second-most-common mechanism
L (Lubrication)	10	9%	Lubrication failures are disproportionately severe (highest severity scores)
T (Time/Temperature)	7	6%	Thermal effects — including thermal bow and Weibull aging
R (Reactive/EM)	6	5%	Motor-specific electrical faults
M (Material/Man)	5	4%	Installation and design errors
E (Environment)	4	3%	Contamination and corrosion

FRETTLSM Cross-Reference by Component Type

This table shows which specific rules fall under each FRETTLSM category for every component type:

Component	F	R	E	T	L	S	M
AFB	AFB01,06,07,08,09,10	AFB12	AFB05,15	AFB14	AFB03,04,16	AFB11,13,15	AFB02,13
JB	JB02,03,05,08,09,12	—	JB04	JB07,11	JB01,02,10	JB06,08	JB11
TPJB	TPJB01,02,07,09,11,12	—	TPJB10	TPJB06	TPJB04	TPJB02,03,05A,05B,08,10	—
COUP	COUP01,02,03A,03B,04,06	—	—	COUP07	—	COUP05,08	—
AC	AC01,04,05	AC02,03,06,08	—	—	—	AC07	AC09
DC	DC03,06	DC04	—	—	—	DC01,02,05,07	—
GEAR	GEAR04,06,08	—	—	—	GEAR03	GEAR01,02,03,05,07,09,10	GEAR06,07
FND	FND01,02,03,04,05,06,09	—	—	FND07	—	FND04,05,08,10	—
FL	Most (F dominant)	—	FL007	—	—	FL008,009	—
B	B01,03,05	—	—	—	—	B02,04	—
C	C02,04	—	—	—	—	C01,03	—
S (Shafts)	S01,02,03,06	—	—	S07	—	S04,05,08,09	S08

Equipment-Specific Factor Templates

The 88 factors are not all relevant to every machine type. FRETTLSM provides asset class templates that activate the relevant subset with machine-specific weights:

Asset Class	Active Factors (of 88)	Top Initiators	Top Accelerators	Key Retarders
Centrifugal Pump	41	F002 (misalignment), S008 (cavitation), L007 (under-lube), S002 (pitting), L002 (grease depletion)	L003 (water in oil), R011 (erosion-corrosion), TT001 (thermal cycling), L004 (particle contamination), E003 (dust)	Proper alignment, adequate NPSH margin, correct lube schedule, seal flush plan, vibration monitoring, minimum flow protection
Gearbox	40	GEAR03 (scuffing), GEAR02 (pitting), F002 (misalignment), L007 (under-lube), S003 (spalling)	L003 (water), L004 (particles), TT001 (cycling), E003 (dust)	Correct oil viscosity, alignment, tooth profile verification, oil filtration
Electric Motor	38	R005 (winding fault), F001 (imbalance), F002 (misalignment), T008 (insulation aging)	E001 (high ambient temp), TT002 (hot spots), L003 (water)	Adequate cooling, voltage balance, proper bearing lubrication
Fan	35	F001 (imbalance), F005 (resonance), F002 (misalignment), S008 (erosion)	E003 (dust), TT001 (cycling), L003 (water)	Balance program, foundation stiffness, ductwork design
Compressor	42	F009 (axial thrust), S008 (erosion), F002 (misalignment), L007 (under-lube), FL015 (surge)	TT001 (cycling), L003 (water), E001 (high ambient)	Surge control, proper sealing, adequate intercooling

Complete Rule Inventory

RAPID AI’s rule system spans four pipeline layers corresponding to ISO 13374 information processing levels. Every rule belongs to exactly one layer:

Layer	Purpose	Pipeline Stage	ISO 13374 Level	Rule Count
GUARD	Data Quality Gate	Module A entry	L2	16
SENSE	Signal Intelligence	Modules A + B	L2-L3	275+
FUSE	Fusion and Prognostics	Modules C + D	L4-L5	66
ACT	Advisory and Action	Module E	L6	46
Total				454

GUARD Rules (DG001-DG019) — Data Quality Gate

Every request must pass through GUARD before entering the engine. The quality score from GUARD propagates through all downstream calculations via S_eff = S_fusion x Q_data and C_final = Q_data x (1 - product(1 - C_i)).

Canonical reference: See Chapter 6 for the authoritative confidence propagation formula.

Quality Score Formula:

quality_score = product(penalty_i)   for all triggered soft rules

quality_score	Status	Pipeline Effect
>= 0.80	pass	Normal confidence — proceed
0.50-0.80	warn	Proceed with reduced confidence
< 0.50	fail	Results unreliable
0.0 (hard block)	block	Pipeline aborts

Hard Block Rules (pipeline aborts):

Rule ID	Canonical ID	Condition	Action	Status
DG_001	GUARD.DG.001	Missing asset_id or timestamp	Block	Spec
DG_002	GUARD.DG.002	len(values) < 256 samples	Block	Implemented
DG_005	GUARD.DG.005	Unit not in allowed set for signal type	Block	Implemented

Soft Penalty Rules (reduce quality_score):

Rule ID	Canonical ID	Condition	Penalty	Effect	Status
DG_003	GUARD.DG.003	nan_fraction > 0.01	x0.6	NaN contamination	Implemented
DG_004	GUARD.DG.004	clip_fraction > 0.01	x0.5	Sensor saturation / clipping	Spec
DG_006	GUARD.DG.006	Sampling rate not in allowed set	x0.7	Suspect acquisition	Spec
DG_007	GUARD.DG.007	Flatline (std of diff < 1e-6)	x0.4	Dead sensor	Implemented
DG_009	GUARD.DG.009	Z-score outlier fraction > 0.02	x0.9	Burst noise	Spec
DG_010	GUARD.DG.010	kurtosis > 8 OR crest_factor > 6	x0.6	Spike burst	Spec
DG_013	GUARD.DG.013	RPM missing from context	x0.85	Reduced feature set	Spec
DG_016	GUARD.DG.016	Magnet mount + RMS > 10	x0.8	Mount slip risk	Spec

Reserved Rules (not yet implemented):

Rule ID	Canonical ID	Intended Purpose	Status
DG_008	GUARD.DG.008	DC offset detection	Reserved — v2.0
DG_011	GUARD.DG.011	Aliasing detection (Fs < 2.5x Fmax)	Reserved — v2.0
DG_012	GUARD.DG.012	Phase reversal check (multi-channel)	Reserved — multi-sensor
DG_014	GUARD.DG.014	Signal duration mismatch	Reserved — v2.0
DG_015	GUARD.DG.015	Cross-channel synchronization	Reserved — multi-sensor

SENSE Rules — Component Fault Initiators (119 Rules)

These are physics-based fault detection rules organized by component type. Each rule identifies a specific fault initiator through vibration directional ratios, impulse metrics, and process conditions. Rules are evaluated in priority order — first match wins per component.

Rule Count by Component Type:

Component Type	Key	Rule Count	FRETTLSM Distribution	P-F Range
Anti-Friction Bearings	afb	16	F:4, L:4, S:3, E:2, R:1, T:1, M:1	P1-P4
Journal Bearings	journal	12	L:4, F:4, S:2, T:1, M:1	P2-P5
Tilting Pad Journal Bearings	tpjb	13	F:4, S:4, L:2, T:2, E:1	P2-P5
Couplings	coupling	9	F:5, S:3, T:1	P2-P3
AC Motors	ac_motor	9	F:3, R:4, S:1, M:1	P2-P3
DC Motors	dc_motor	7	S:4, F:2, R:1	P2-P4
Gearboxes	gears	10	S:6, F:3, M:1	P2-P5
Foundation / Structure	foundation	10	F:5, S:4, T:1	P2-P4
Fluid Flow / Pumps	fluid_flow	15	F:12, S:2, E:1	P2-P5
Belt Drives	belts	5	F:3, S:2	P2-P4
Chain Drives	chains	4	S:2, F:2	P2-P3
Shafts	shafts	9	F:3, S:4, T:1, M:1	P2-P5
Total		119	F:52, S:36, L:10, R:6, T:7, E:4, M:5

AFB — Anti-Friction Bearings (16 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions (AND unless noted)	P-F
AFB01	Incorrect Preload — High	F	0.60	A_H_ratio >= 1.2 AND V_H_ratio < 0.8	P2
AFB02	Wrong Clearance	M	0.50	H_V_ratio >= 1.6 OR H_V_ratio < 0.6	P2
AFB03	Lubrication Starvation	L	0.80	H_V_ratio in [0.9, 1.1] AND kurtosis >= 5.0	P3
AFB04	Wrong Lubricant Viscosity	L	0.50	H_V_ratio in [0.9, 1.1] AND temperature >= 50C	P2
AFB05	Contamination	E	0.60	kurtosis >= 5.0 AND crest_factor >= 2.0	P3
AFB06	Shaft Imbalance	F	0.50	H_V_ratio in [1.2, 2.0] AND A_H_ratio < 0.3	P2
AFB07	Coupling Misalignment	F	0.60	A_H_ratio >= 1.3 AND A >= 0.5 mm/s	P2
AFB08	Overloading	F	0.60	H_V_ratio >= 1.1 AND temperature >= 65C	P3
AFB09	Resonance	F	0.80	H >= 3.0 mm/s AND H_V_ratio >= 3.0	P4
AFB10	Soft Foot	F	0.50	V_H_ratio >= 1.4 AND A_V_ratio < 0.3	P2
AFB11	Poor Surface Finish	S	0.50	kurtosis >= 5.0 AND crest_factor >= 2.5	P1
AFB12	VFD Bearing Currents	R	0.70	crest_factor >= 3.0 AND kurtosis >= 6.0	P3
AFB13	Incorrect Mounting	M	0.50	H_V_ratio >= 1.5 AND A_H_ratio < 0.5	P2
AFB14	Thermal Expansion Effects	T	0.50	A_H_ratio >= 1.1 AND temperature >= 60C	P2
AFB15	False Brinelling	E	0.40	H < 0.2 AND V < 0.2 AND kurtosis >= 7.0	P2
AFB16	Early Micro-Slip	L	0.40	H_V_ratio in [0.9, 1.1] AND temperature >= 50C	P1

JB — Journal Bearings (12 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
JB01	Low Oil Film Thickness	L	0.60	H_V_ratio >= 1.3 AND temperature >= 70C	P3
JB02	Oil Starvation	L	0.90	H_V_ratio in [0.9, 1.1] AND kurtosis >= 6.0	P4
JB03	Rotor Instability / Oil Whirl	F	0.70	H_V_ratio in [0.9, 1.1]	P4
JB04	Oil Contamination	E	0.60	kurtosis >= 4.5 AND temperature >= 60C	P2
JB05	Shaft Misalignment	F	0.60	A_H_ratio >= 1.2	P2
JB06	Excessive Clearance	S	0.60	H_V_ratio >= 2.0	P3
JB07	Thermal Distortion	T	0.50	temperature >= 75C AND H_V_ratio >= 1.3	P2
JB08	Rotor Rubbing	F	0.90	A_H_ratio >= 0.7 AND kurtosis >= 7.0	P5
JB09	Rotor Bow	F	0.60	H_V_ratio in [1.2, 1.8]	P3
JB10	Shaft Settling	L	0.50	V_H_ratio >= 1.5 AND kurtosis >= 4.0	P2
JB11	Assembly Error	M	0.70	H_V_ratio in [0.9, 1.1] AND temperature >= 80C	P2
JB12	Heavy Rotor Load	F	0.50	H_V_ratio >= 1.4	P2

TPJB — Tilting Pad Journal Bearings (13 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
TPJB01	Pad Misalignment	F	0.60	H_V_ratio in [0.9, 1.1]	P3
TPJB02	Preload Loss	S	0.70	H_V_ratio >= 1.3 AND A_H_ratio < 0.3	P4
TPJB03	Pad Pivot Wear	S	0.60	kurtosis >= 4.0 AND crest_factor >= 2.5	P3
TPJB04	Oil Starvation	L	0.90	temperature >= 75C AND H_V_ratio in [0.9, 1.1]	P5
TPJB05A	Babbitt Fatigue (Thermal)	S	0.80	kurtosis >= 5.0 AND temperature >= 65C	P4
TPJB05B	Babbitt Fatigue (Mechanical)	S	0.80	kurtosis >= 5.0 AND temperature >= 65C	P4
TPJB06	Thermal Distortion	T	0.50	temperature >= 70C AND A_H_ratio >= 0.7 AND V >= 1.5	P2
TPJB07	Shaft Misalignment	F	0.60	A_H_ratio >= 1.2	P2
TPJB08	Excessive Clearance	S	0.60	H_V_ratio >= 2.0	P3
TPJB09	Oil Whip (Severe)	F	0.90	H_V_ratio in [0.9, 1.1] AND kurtosis >= 6.0	P5
TPJB10	Contamination	E	0.60	kurtosis >= 5.0 AND crest_factor >= 3.0	P3
TPJB11	Overload	F	0.70	H_V_ratio >= 1.4 AND temperature >= 70C	P3
TPJB12	Rotor Rub	F	0.90	A_H_ratio >= 0.7 AND kurtosis >= 7.0	P5

COUP — Couplings (9 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
COUP01	Parallel Misalignment	F	0.60	H_V_ratio >= 1.3 AND A_H_ratio < 0.2	P2
COUP02	Angular Misalignment	F	0.60	A_H_ratio >= 1.3	P2
COUP03A	Coupling Unbalance (Static)	F	0.50	kurtosis >= 4.0 AND crest_factor >= 2.0	P3
COUP03B	Coupling Unbalance (Dynamic)	F	0.50	kurtosis >= 4.0 AND crest_factor >= 2.0	P3
COUP04	Combined Misalignment + Unbalance	F	0.60	H_V_ratio >= 1.2 AND A_H_ratio in [0.3, 0.8]	P3
COUP05	Coupling Backlash / Wear	S	0.50	A_H_ratio >= 0.8 AND kurtosis >= 4.0	P3
COUP06	Torsional Vibration via Coupling	F	0.60	H_V_ratio >= 1.1 AND kurtosis >= 3.5	P3
COUP07	Thermal Growth Misalignment	T	0.50	A_H_ratio >= 1.0 AND temperature >= 55C	P2
COUP08	Disc Pack Fatigue	S	0.70	crest_factor >= 3.0 AND A_H_ratio >= 0.6	P3

AC — AC Motors (9 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
AC01	Rotor Imbalance	F	0.50	H_V_ratio in [1.2, 2.5] AND A_H_ratio < 0.3	P2
AC02	Rotor Bar Crack	R	0.60	H_V_ratio in [1.1, 2.0]	P3
AC03	Air-Gap Eccentricity	R	0.60	H_V_ratio >= 2.0 AND A_H_ratio < 0.2	P3
AC04	Soft Foot	F	0.50	V_H_ratio >= 1.5 AND A_H_ratio < 0.5	P2
AC05	Misalignment	F	0.60	A_H_ratio >= 1.3	P2
AC06	Stator Eccentricity	R	0.50	H_V_ratio >= 1.5 AND A_H_ratio < 0.3	P2
AC07	Motor Bearing Fault	S	0.70	H_V_ratio in [0.9, 1.1]	P3
AC08	VFD Harmonic Excitation	R	0.50	H_V_ratio >= 1.0 AND kurtosis >= 4.0 AND crest_factor >= 2.5	P3
AC09	Loose Rotor on Shaft	M	0.60	H_V_ratio >= 1.2 AND kurtosis >= 3.5	P2

DC — DC Motors (7 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
DC01	Commutator Roughness	S	0.50	kurtosis >= 4.0 AND crest_factor >= 2.5	P2
DC02	Brush Wear	S	0.60	kurtosis >= 5.0 AND H_V_ratio in [0.9, 1.1]	P3
DC03	Armature Imbalance	F	0.50	H_V_ratio >= 1.2 AND A_H_ratio < 0.3	P2
DC04	Field Winding Fault	R	0.70	H_V_ratio >= 2.0	P4
DC05	Bearing Fault	S	0.70	kurtosis >= 6.0 AND crest_factor >= 3.0	P3
DC06	Misalignment	F	0.60	A_H_ratio >= 1.3	P2
DC07	Commutator Eccentricity	S	0.60	H_V_ratio >= 1.5 AND kurtosis >= 4.0	P3

GEAR — Gearboxes (10 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
GEAR01	Uniform Tooth Wear	S	0.50	H_V_ratio >= 1.2 AND H >= 2.0 mm/s	P2
GEAR02	Localized Pitting	S	0.60	kurtosis >= 4.0	P3
GEAR03	Tooth Root Crack	S	0.75	kurtosis >= 4.5 AND crest_factor >= 2.5	P4
GEAR04	Gear Misalignment	F	0.60	A_H_ratio >= 1.2 AND H_V_ratio >= 1.2	P2
GEAR05	Backlash Excess	S	0.50	kurtosis >= 4.5 AND H_V_ratio >= 1.3	P3
GEAR06	Hunting Tooth Frequency	M	0.40	H >= 1.5 mm/s AND H_V_ratio >= 1.1	P3
GEAR07	Ghost Frequency	M	0.30	H >= 2.5 mm/s AND kurtosis < 3.5	P3
GEAR08	Gear Natural Frequency Excitation	F	0.70	H >= 4.0 mm/s AND H_V_ratio >= 2.0	P4
GEAR09	Tooth Profile Error	S	0.50	H_V_ratio >= 1.3 AND kurtosis >= 3.5	P3
GEAR10	Tooth Chipping / Breakage	S	0.80	kurtosis >= 5.0	P5

FND — Foundation / Structure (10 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
FND01	Soft Foot	F	0.50	V_H_ratio >= 1.4	P2
FND02	Base Looseness	F	0.60	H_V_ratio >= 1.6	P3
FND03	Structural Resonance	F	0.70	H_V_ratio >= 3.0	P4
FND04	Anchor Bolt Fatigue	S	0.50	H_V_ratio in [1.4, 2.5]	P2
FND05	Grout Deterioration	S	0.50	H_V_ratio >= 1.3 AND V >= 1.0 mm/s	P3
FND06	Piping Strain	F	0.50	A_H_ratio >= 0.8 AND V_H_ratio >= 1.2	P2
FND07	Thermal Expansion of Structure	T	0.40	H_V_ratio >= 1.3 AND temperature >= 50C	P2
FND08	Structural Crack	S	0.80	H_V_ratio >= 2.0 AND H >= 3.0 mm/s	P4
FND09	Baseplate Resonance	F	0.60	H_V_ratio >= 2.5 AND H >= 2.5 mm/s	P3
FND10	Pedestal Crack	S	0.80	H >= 3.5 mm/s AND kurtosis >= 4.0	P4

FL — Fluid Flow / Pumps (15 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
FL001	Micro-bubble Formation	F	0.50	H_V_ratio in [0.9, 1.2]	P2
FL002	Continuous Cavitation	F	0.80	H >= 1.3 AND V >= 1.3 AND crest_factor >= 3.0	P4
FL003	Discharge Recirculation	F	0.60	H_V_ratio in [0.8, 1.3] AND kurtosis >= 3.5	P3
FL004	Suction Recirculation	F	0.60	kurtosis >= 4.0 AND H_V_ratio in [0.9, 1.1]	P3
FL005	Vane Pass Frequency Excitation	F	0.60	H >= 2.0 AND V >= 1.5 mm/s	P3
FL006	Water Hammer	F	0.90	crest_factor >= 4.5	P5
FL007	Air Entrainment	E	0.50	kurtosis >= 3.5 AND crest_factor >= 2.5	P2
FL008	Impeller Wear / Erosion	S	0.50	H_V_ratio in [1.2, 2.0]	P3
FL009	Seal Leakage / Failure	S	0.60	A_H_ratio >= 0.7 AND kurtosis >= 3.0	P3
FL010	Hydraulic Instability	F	0.60	H_V_ratio in [0.9, 1.2] AND crest_factor >= 2.5	P3
FL011	Dead-heading	F	0.90	temperature >= 75C AND H_V_ratio in [0.9, 1.1]	P5
FL012	BEP Deviation — Low Flow	F	0.50	H_V_ratio >= 1.3 AND V >= 1.5 mm/s	P3
FL013	BEP Deviation — High Flow	F	0.50	H >= 2.0 AND A_H_ratio >= 0.5	P3
FL014	Bearing Overload from Thrust	F	0.70	A_H_ratio >= 1.0 AND temperature >= 65C	P3
FL015	Surge / Stall in Compressor	F	0.90	A_H_ratio >= 1.0 AND crest_factor >= 4.0	P5

B — Belt Drives (5 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
B01	Belt Tension Incorrect	F	0.50	H_V_ratio >= 1.3	P2
B02	Belt Wear / Cracking	S	0.60	kurtosis >= 4.0 AND crest_factor >= 2.0	P3
B03	Sheave Misalignment	F	0.50	A_H_ratio >= 1.2	P2
B04	Sheave Wear	S	0.50	H_V_ratio >= 1.4 AND kurtosis >= 3.5	P3
B05	Belt Resonance	F	0.70	H >= 2.0 mm/s AND H_V_ratio >= 2.5	P4

C — Chain Drives (4 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
C01	Chain Wear / Elongation	S	0.50	kurtosis >= 4.0	P3
C02	Chain Tension Incorrect	F	0.50	H_V_ratio >= 1.3 AND kurtosis >= 3.5	P2
C03	Sprocket Wear	S	0.60	kurtosis >= 5.0 AND crest_factor >= 2.5	P3
C04	Chain Misalignment	F	0.50	A_H_ratio >= 1.2	P2

S — Shafts (9 Rules):

Rule ID	Initiator	FRETTLSM	Severity	Conditions	P-F
S01	Uneven Mass Distribution	F	0.50	H_V_ratio >= 1.2 AND A_H_ratio < 0.3	P2
S02	Shaft Bow	F	0.60	H_V_ratio in [1.1, 1.8] AND A_H_ratio < 0.4	P3
S03	Coupling Face Misalignment	F	0.60	A_H_ratio >= 1.3 AND A_V_ratio >= 1.3	P2
S04	Keyway Damage	S	0.50	kurtosis >= 4.0 AND H_V_ratio >= 1.1	P3
S05	Shaft Crack	S	0.95	A_H_ratio >= 0.8 AND kurtosis >= 4.0	P5
S06	Torsional Vibration	F	0.60	H_V_ratio >= 1.0 AND kurtosis >= 3.5 AND crest_factor >= 2.0	P3
S07	Thermal Bow	T	0.60	H_V_ratio >= 1.2 AND temperature >= 65C	P3
S08	Coupling Hub Runout	M	0.40	H_V_ratio >= 1.2 AND A_H_ratio in [0.2, 0.6]	P2
S09	Shaft Surface Damage	S	0.50	kurtosis >= 4.5 AND H_V_ratio >= 1.0	P3

SENSE Rules — Signal Feature Rules (SF001-SF051)

Cross-component rules evaluated for every analysis, not filtered by component type. These detect fault patterns that span across equipment types.

Rule ID Range	Count	Group	Description
SF001-SF008	8	Imbalance and Resonance	Classic imbalance, combined faults, sub-sync, super-sync, broadband, sidebands, impact series, resonance dwell
SF009-SF017	9	Bearing Life Stages	Early inner race, outer race, BPFO+, BPFI+, BSF, ball spin, FTF, distributed wear, severe spall
SF018-SF025	8	Spectral Entropy and Stability	Stable healthy, early degradation, spectral disorder, chaotic multi-fault, rapid deterioration, entropy spike, plateau, recovery
SF026-SF034	9	Process and Load Correlation	Cavitation onset, advanced cavitation, pump off-BEP (low/high), load-proportional, intermittent, transient startup, coast-down anomaly, motor slip
SF035-SF042	8	Waveform and Statistical	Kurtosis onset/alarm, crest factor moderate/severe, negative kurtosis, bimodal distribution, periodic/random impulse
SF043-SF050	8	Trend and Rate of Change	Stable healthy trend, slow drift, accelerating early/severe, step change, oscillating trend, recovery trend, post-maintenance Waddington

SENSE Rules — SEDL Entropy State Rules

The thermodynamic backbone of RAPID AI. Computes entropy metrics that detect degradation before amplitude changes become visible.

Rule	Condition	State	Severity
SR05	SI <= 0.40	Critical_Instability	alarm
SR04	SE >= 0.65 AND (TE >= 0.60 OR DE >= 0.60)	Chaotic	warning
SR03	dSE_dt >= 0.02 AND SI < 0.60	Destabilizing	warning
SR02	SE > 0.35 AND SE < 0.65 AND dSE_dt < 0.02	Drifting	watch
SR01	SE <= 0.35 AND TE < 0.50 AND SI >= 0.70	Stable	normal

SENSE Rules — Slope Intelligence

Class	Condition	Severity
Step	max_jump > 0.5 (sudden shift)	0.80
Chaotic	volatility > 0.3 AND \|slope\| < 0.02	0.30 (investigate, not alarm)
Accelerating	\|slope\| > 0.05 AND \|slope_change\| > 0.02	min(1.0, \|slope\|x10 + \|slope_change\|x5)
Drift	\|slope\| > 0.02 (steady rise)	min(0.70, \|slope\|x8)
Stable	Otherwise	max(0.0, \|slope\|x3)

SENSE Rules — Signal Group Rules (79 Rules, Documented)

Group	Rule Range	Count	Description	Status
SENSE.VT	VT001-VT005	5	Velocity amplitude trend classification	Spec
SENSE.AT	AT001-AT006	6	Acceleration amplitude trend classification	Spec
SENSE.DT	DT001-DT005	5	Displacement amplitude trend classification	Spec
SENSE.FB	FB001-FB015	15	FFT band energy rules	Spec
SENSE.HF	HF001-HF007	7	High-frequency demodulation rules	Spec
SENSE.IM	IM001-IM006	6	Impulse metric rules	Spec
SENSE.EA	EA001-EA007	7	Envelope analysis rules	Spec
SENSE.ET	ET001-ET009	9	Electrical signature rules	Spec
SENSE.TM	TM001-TM008	8	Temperature trend rules	Spec

FUSE Rules — Block Scoring (BSR001-BSR007)

Each component block gets scored based on its Module B (initiator match) and Module B.2 (trend) outputs. Rules are evaluated in priority order — first match wins.

Rule ID	Canonical ID	Condition	Block Score	Block State	Priority
BSR007	FUSE.BS.007	B_match_score >= 0.90	0.90	critical	1
BSR001	FUSE.BS.001	Trend = Accelerating AND confidence >= 0.70	0.85	unstable	2
BSR003	FUSE.BS.003	Trend = Step AND confidence >= 0.70	0.80	unstable	2
BSR002	FUSE.BS.002	Trend = Drift AND confidence >= 0.60	0.65	degrading	4
BSR005	FUSE.BS.005	B_match >= 0.70 AND (Trend = Stable OR confidence < 0.50)	0.55	watch	3
BSR004	FUSE.BS.004	Trend = Chaotic AND process_correlation >= 0.70	0.35	process-driven	5
BSR006	FUSE.BS.006	B_match < 0.30 AND Trend = Stable	0.15	healthy	6

SSI Formula:

SSI = sum(weight_i x block_score_i)   for all blocks in machine profile, clamped to [0, 1]

Gating Rule: If stability_state == ‘Critical_Instability’, SSI cannot drop below 0.70 regardless of block scores.

SSI to System State Mapping:

SSI Range	System State	Severity Level
< 0.30	stable / healthy	normal
0.30-0.60	degrading / watch	watch
0.60-0.80	unstable / warning	warning
>= 0.80	critical / alarm	alarm

Machine Profiles (component block weights):

Block	Pump Train	Gearbox Train	Fan Train
fluid_flow	0.30 (req)	—	0.30 (req)
gears	—	0.35 (req)	—
afb (bearings)	0.25 (req)	0.20 (req)	0.25 (req)
foundation	0.15 (req)	0.15 (req)	0.15 (req)
ac_motor	0.10 (req)	0.10 (req)	0.15 (req)
coupling	0.10	0.10	0.05
shafts	0.10 (req)	0.10 (req)	0.10 (req)

FUSE Rules — Health Staging and RUL

Module D maps SSI + SSI_slope into a health stage with slope-based escalation and remaining useful life bands.

Health Stages:

Stage ID	Condition	Stage Name	RUL Band	Min Days	Max Days
HS01	SSI < 0.30	Healthy	Long	180	—
HS02	SSI >= 0.30 AND SSI < 0.60	Degrading	Medium	30	180
HS03	SSI >= 0.60 AND SSI < 0.80	Unstable	Short	7	30
HS04	SSI >= 0.80	Critical	Immediate	0	7

Slope Escalation:

Current Stage	Slope Condition	Escalates To
Healthy	SSI_slope > 0.02	Degrading
Degrading	SSI_slope > 0.05	Unstable

Escalation Rules:

Rule ID	Condition	Escalation Level	Recommended Action
ER04	Stage = Critical	Level_3	Immediate intervention required
ER05	SSI >= 0.80 (override)	Level_3	Immediate intervention required
ER03	Stage = Unstable	Level_2	Prepare intervention plan
ER02	Stage = Degrading AND SSI_slope > 0.03	Level_1	Schedule inspection
ER01	Stage = Healthy	Level_0	Continue monitoring

ACT Rules — Maintenance Action Selection

Module E converts the health picture into concrete maintenance decisions.

Priority Score Formula:

Canonical reference: See Chapter 26 for the authoritative priority formula.

P = 100 x (0.45*S + 0.25*C + 0.20*K + 0.10*U) x M_safe x R_sp x R_mp

Where S = severity, C = confidence, K = asset criticality, U = time urgency, M_safe = safety multiplier (1.5 if safety consequence), R_sp = spares readiness (0.7 if unavailable), R_mp = manpower readiness (0.7 if constrained).

Priority Windows:

Priority Range	Window	Meaning
P >= 85	Immediate	Stop/inspect now or within shift
70 <= P < 85	24 hours	Execute within 24 hours
50 <= P < 70	7 days	Plan within week
P < 50	Next shutdown	Bundle into planned outage

Action Catalog (ACT001-ACT008):

Action ID	Title	Justification	Verification
ACT001	Vibration re-measure (confirmation run)	Confirm trend before committing to intervention	Compare new RMS to previous within +/-10%
ACT002	Bearing lubrication / grease replenishment	HF or temperature rise indicates lubrication deficit	HF amplitude and temperature return to baseline within 24h
ACT003	Alignment check (laser / dial indicator)	Axial dominance and coupling signature indicate misalignment	Alignment report within tolerance per OEM spec
ACT004	Balance correction (single / dual plane)	1x dominance with horizontal preference indicates imbalance	1x amplitude reduced by >= 50% post-correction
ACT005	Bearing replacement (scheduled)	BPFO/BPFI signatures with acceleration confirms defect	Post-replacement vibration within acceptance limits
ACT006	Foundation tightening / soft foot correction	V/H ratio and looseness indicate structural issues	Phase stability and reduced looseness harmonics
ACT007	Process investigation (not machine fault)	Trend correlates with process variable, not degradation	Vibration returns to normal with process stabilization
ACT008	Emergency shutdown / trip recommendation	Critical SSI with accelerating trend, imminent failure risk	Machine isolated, inspection completed before restart

Diagnosis to Action Mapping:

Keyword in Diagnosis	Action IDs Selected
imbalance / unbalance	ACT004, ACT001
misalignment	ACT003, ACT001
bearing	ACT002, ACT005
lubrication	ACT002
looseness	ACT006, ACT001
foundation	ACT006
process	ACT007
critical / shutdown	ACT008

Selection Logic: (1) If P >= 85, always include ACT008. (2) Match diagnosis keywords to append corresponding actions. (3) If no keywords match, default to ACT001 (confirmation run). (4) Deduplicate while preserving priority order.

System State Action Playbook:

System State	Default Action	Operator Message
stable	monitor	System stable. Continue routine monitoring.
degrading	alert	Plan inspection within schedule window.
unstable	intervene	Reduce load/speed if possible. Inspect soon.
critical	shutdown_or_trip	Stop if safe. Inspect immediately.
process-driven	investigate_process	Check operating conditions. Re-test after stabilization.

Shared Fault Patterns (Cross-Component)

Seven fault patterns repeat across multiple component types with identical physics:

Universal Fault	FRETTLSM	Components Affected	Rule IDs
Mass Imbalance	F	AFB, AC, DC, S, COUP, FND	AFB06, AC01, DC03, S01, COUP04, FND04
Angular Misalignment	F	AFB, JB, TPJB, COUP, AC, DC, B, C, S, GEAR	AFB07, JB05, TPJB07, COUP02, AC05, DC06, B03, C04, S03, GEAR04
Soft Foot / Looseness	F	AFB, AC, FND	AFB10, AC04, FND01
Structural Resonance	F	AFB, FND, B	AFB09, FND03, B05
Contamination / Impact	E/S	AFB, JB, TPJB, B, C	AFB05, JB04, TPJB10, B02, C01
Overload + Thermal	F/T	AFB, JB, TPJB	AFB08, JB01, TPJB11
Lubrication Starvation	L	AFB, JB, TPJB, GEAR	AFB03, JB02, TPJB04, GEAR01

Critical Safety Rules (Severity >= 0.80)

These rules have safety-level or catastrophic consequences and should trigger immediate operator notification regardless of pipeline confidence level:

Rule ID	Initiator	Severity	RCM Consequence	P-F
S05	Shaft Crack	0.95	safety	P5
TPJB04	Oil Starvation	0.90	safety	P5
TPJB09	Oil Whip (Severe)	0.90	safety	P5
TPJB12	Rotor Rub	0.90	safety	P5
FL006	Water Hammer	0.90	safety	P5
FL011	Dead-heading	0.90	safety	P5
FL015	Surge / Stall	0.90	safety	P5
JB02	Oil Starvation	0.90	safety	P4
JB08	Rotor Rubbing	0.90	safety	P5
AFB03	Lubrication Starvation	0.80	operational	P3
AFB09	Resonance	0.80	safety	P4
TPJB05A/B	Babbitt Fatigue	0.80	safety	P4
FL002	Continuous Cavitation	0.80	environmental	P4
GEAR10	Tooth Breakage	0.80	safety	P5
FND08	Structural Crack	0.80	safety	P4
FND10	Pedestal Crack	0.80	safety	P4

Rule Evaluation Mechanics

How Rules Are Structured

Every rule in the RAPID AI system conforms to a canonical schema that captures not just detection logic but the full diagnostic chain — from physics to consequence to maintenance action. The schema uses a {LAYER}.{GROUP}.{SEQ} naming convention (e.g., SENSE.CF.AFB03) while preserving legacy IDs for backward compatibility.

Canonical Rule Schema:

Field	Type	Required	Description
`id`	string	Yes	Canonical ID: `{LAYER}.{GROUP}.{SEQ}`
`rule_id` / `legacy_id`	string	Yes	Legacy short ID (e.g., `AFB03`)
`version`	int	No	Schema version for migration tracking
`initiator` / `name`	string	Yes	Short name of the fault initiating condition
`diagnosis`	string	Yes	What failure mechanism is occurring physically
`physics`	string	Yes	Full physical explanation of the energy flow disruption
`consequence`	string	Yes	What happens if the fault is left unaddressed
`component_type`	string	Yes	Component key: `afb`, `journal`, `tpjb`, `coupling`, `ac_motor`, `dc_motor`, `gears`, `foundation`, `fluid_flow`, `belts`, `chains`, `shafts`
`frettlsm`	string	Yes	Diagnostic lens category: F, R, E, T, L, S, or M
`conditions`	list	Yes	Primary AND conditions — all must match
`or_conditions`	list	No	Fallback OR conditions — any single match fires at reduced confidence
`supplementary`	list	No	Informational signals from CSV — not evaluated by engine
`severity_base`	float	Yes	Inherent severity 0.0-1.0
`severity_stages`	dict	No	Early/mid/late severity progression (future)
`rcm_consequence`	string	No	safety, environmental, operational, hidden
`rcm_task_type`	string	No	condition_based, scheduled, redesign, run_to_failure
`pf_position`	string	No	P1 (earliest) through P5 (near functional failure)
`cbm_technique`	string	No	vibration_analysis, temperature_monitoring, electrical_analysis, envelope_analysis, process_monitoring
`iso_14224_mechanism`	string	No	ISO 14224 failure mechanism category
`iso_14224_cause`	string	No	ISO 14224 failure cause category
`recommended_actions`	list	No	Specific corrective guidance strings
`tags`	list	No	Semantic search tags
`references`	list	No	External standards references

Example Rule (AFB03 — Lubrication Starvation):

id:            SENSE.CF.AFB03
legacy_id:     AFB03
initiator:     Lubrication Starvation
diagnosis:     "Film collapse; metal-to-metal contact generating HF impacts"
physics:       "Grease has dried out or oil feed is blocked. Without an
               elastohydrodynamic film, rolling elements contact race surfaces
               directly. Each asperity collision generates a micro-impact."
consequence:   "Rapid surface degradation within days to weeks. Bearing
               temperature rises. Without intervention, progresses to spalling,
               cage fracture, and seizure. P-F interval as short as 2-4 weeks."
component_type: afb
frettlsm:     L

conditions:
  - metric: H_V_ratio
    op: ">="
    value: 0.9
  - metric: H_V_ratio
    op: "<="
    value: 1.1
  - metric: kurtosis
    op: ">="
    value: 5.0

supplementary:
  - metric: crest_factor
    op: ">"
    value: 2.0
  - metric: temperature
    op: ">"
    value: 60

severity_base: 0.80
pf_position: P3
rcm_consequence: operational
iso_14224_mechanism: wear
iso_14224_cause: lubrication_failure

recommended_actions:
  - "Check lubrication supply — confirm flow rate to bearing"
  - "Check lubricant quantity — refill grease or check oil level"
  - "Check for supply line blockage or filter obstruction"
  - "Verify lubricant type and viscosity grade"

Rule Evaluation Logic (AND / OR)

The rule engine evaluates in two passes:

Pass 1 (AND): Check ALL primary conditions. Every condition in the conditions list must be satisfied for the rule to fire at full severity.
Pass 2 (OR fallback): If primary conditions fail and or_conditions exist, any single match fires the rule at reduced confidence (0.5x severity_base).

Condition operators: >=, <=, >, <, ==

Complete Metric Inventory

All rules reference one set of computed metrics. The rule engine computes derived ratios before evaluation.

Raw Input Metrics:

Metric Key	Units	Description	Source
H	mm/s RMS	Horizontal velocity — drive-end radial	SignalInput
V	mm/s RMS	Vertical velocity — radial, 90 degrees from H	SignalInput
A	mm/s RMS	Axial velocity — thrust direction	SignalInput
RMS	mm/s	Overall RMS (all axes combined)	Computed in Module A
temperature	C	Bearing or housing temperature	ContextInput
kurtosis	dimensionless	Statistical kurtosis of time waveform	Computed from raw signal
crest_factor	dimensionless	Peak / RMS ratio	Computed from raw signal
current	A	Motor current	ContextInput

Derived Ratio Metrics (computed with epsilon guard to prevent division by zero):

Metric Key	Formula	Dominant Fault Signature	Normal Range
H_V_ratio	H / V	Horizontal dominance: imbalance, misalignment, looseness, resonance	0.8-1.4
V_H_ratio	V / H	Vertical dominance: soft foot, piping strain, vertical resonance	0.7-1.2
A_H_ratio	A / H	Axial dominance: angular misalignment, coupling fault, shaft crack	< 0.5
A_V_ratio	A / V	Axial/vertical ratio: used in combination rules	< 0.5
V_A_ratio	V / A	Vertical/axial ratio: rarely used alone	varies

Physical interpretation of ratio ranges:

H_V_ratio 0.6-1.6: Normal energy distribution in radial plane
H_V_ratio 1.6-2.5: Imbalance or clearance issue
H_V_ratio > 2.5: Looseness or resonance amplification
H_V_ratio < 0.6: Soft foot or vertical stiffness anomaly
A_H_ratio < 0.3: Pure radial fault
A_H_ratio 0.3-0.7: Mixed radial/axial force
A_H_ratio 0.7-1.2: Significant axial content (misalignment, thermal growth)
A_H_ratio > 1.2: Axial dominant (angular misalignment, shaft crack, surge)

Impulse and Statistical Metrics:

Metric	Description	Healthy Range	Fault Signatures
kurtosis	4th statistical moment	2.5-3.5	>4.0: contamination/surface damage; >6.0: VFD currents, oil starvation, rub
crest_factor	Peak / RMS ratio	1.5-2.0	>2.0: contamination; >3.0: EDM pitting, pivot wear; >4.5: water hammer

Spectral / Frequency Band Metrics:

Metric Key	Frequency Range	Fault Detection Application
HF_3_7	3-7 kHz	Early bearing lubrication / surface damage
HF_5_12	5-12 kHz	VFD EDM pitting, surface finish defects
HF_8_15	8-15 kHz	VFD-specific EDM signatures
HF_2_5	2-5 kHz	Early micro-slip detection
LF_0_30	0-30 Hz	Oil whirl (0.42-0.48x running speed)
LF_100_300	100-300 Hz	Foundation looseness, structural modes
MF_100_700	100-700 Hz	Gear mesh frequency range
BB_100_1000	100-1000 Hz	Cavitation, turbulence, air entrainment
VPF_amp	N x RPM Hz	Pump hydraulic performance deviation
RMS_1x	1 x RPM Hz	Imbalance, shaft bow, eccentricity
RMS_2x	2 x RPM Hz	Angular misalignment, coupling looseness

Bearing Defect Frequency Metrics:

Frequency	Formula	Description
BPFO	(n/2) x (1 - d/D x cos a) x f	Ball Pass Frequency Outer race
BPFI	(n/2) x (1 + d/D x cos a) x f	Ball Pass Frequency Inner race
BSF	(D/2d) x (1 - (d/D x cos a)^2) x f	Ball Spin Frequency
FTF	0.5 x (1 - d/D x cos a) x f	Fundamental Train (cage) Frequency

Where: n = number of balls, d = ball diameter (mm), D = pitch diameter (mm), a = contact angle, f = shaft frequency (Hz = RPM/60). Matching uses +/-3% tolerance at harmonics 1, 2, and 3.

RCM Cross-Reference

RAPID AI rules answer five of the seven RCM questions defined by SAE JA1011/JA1012:

RCM Question	RAPID AI Answer	Rule Layer
1. What are the functions?	Asset hierarchy (equipment to measurement point)	Schema
2. What are the functional failures?	Severity levels: normal, watch, warning, alarm	ACT.SV
3. What causes each failure?	119 component fault rules across 12 types	SENSE.CF
4. What happens when each failure occurs?	diagnosis + consequence fields per rule	SENSE.CF
5. In what way does each failure matter?	RCM consequence categories per rule	rcm.consequence
6. What can be done to prevent?	Maintenance action catalog (ACT001-ACT008)	ACT.MA
7. What if no suitable task?	Redesign recommendation or run-to-failure	rcm.task_type

RCM Consequence Categories:

Category	Definition	Example Rules
safety	Failure threatens human safety or structural integrity	S05, TPJB09, TPJB12, FL006
environmental	Failure causes environmental harm or fluid release	FL009, FL002, FC011
operational	Failure stops or degrades production output	Most AFB, JB, GEAR, AC, DC, FND rules
hidden	Failure not detectable during normal operation	AFB16, FND06, JB10

ISO 14224 Cross-Reference

Every rule maps to ISO 14224 failure mechanisms and causes:

ISO 14224 Mechanism	FRETTLSM	Example Rules
Fatigue	S, T	AFB01, AFB09, TPJB05, S05
Wear	S, L	AFB03, JB02, GEAR01, GEAR02, C01
Corrosion	E	FC011, FC014
Erosion	F, E	FL001, FL002, FL006
Deformation	F, T	AFB08, TPJB06, S07
Vibration	F	AFB06, AFB09, AC01, FND03, B05
Overheating	T, L	AFB04, JB01, TPJB04, COUP07
Contamination	E, L	AFB05, JB04, TPJB10, AFB11
Electrical	R	AC03, AC07, DC04, DC05
Clearance change	M, T	AFB02, JB06, TPJB08, GEAR04
Material failure	M	S05, GEAR10, DC04
Blockage / leakage	F	FL007, FL008, FL009

P-F Interval Positions

The P-F curve represents degradation from potential failure (P) to functional failure (F). Every rule fires at a specific position on this curve:

Position	Description	Detection Window	Example Rules
P1	Earliest detectable change	12-24 months before F	AFB16 (micro-slip), AFB11 (surface finish), SEDL dSE/dt onset
P2	Measurable trend	3-12 months before F	AFB01 (preload), AFB06 (imbalance), most COUP rules
P3	Clear symptom	1-3 months before F	AFB03 (lube starvation), AFB05 (contamination), GEAR02 (pitting)
P4	Impending failure	1-4 weeks before F	AFB09 (resonance), JB07 (thermal bow), TPJB05 (babbitt fatigue)
P5	Near functional failure	Days to hours before F	S05 (shaft crack), TPJB04 (oil starvation), FL006 (water hammer)

Key Formulas Reference

Formula	Purpose	Source
`S_eff = S_fusion x Q_data`	Effective severity (data-quality-adjusted)	Module C output
`C_final = Q_data x (1 - (1-C_rules)(1-C_trend)(1-C_entropy))`	Confidence propagation (union of evidence)	Module B output
`SI = 1 - (0.5SE + 0.3TE + 0.2*DE)`	SEDL Stability Index	Module B.3
`SSI = sum(weight_i x block_score_i)`	System Stability Index	Module C
`P = 100 x (0.45S + 0.25C + 0.20K + 0.10U) x M_safe x R_sp x R_mp`	Maintenance priority score	Module E
`beta_adj = beta_base x (1 + alpha_severity x S_eff)`	Condition-adjusted Weibull shape	Module D
`eta_adj = eta_base x (1 - gamma_degradation x SSI)`	Condition-adjusted Weibull scale	Module D
`RUL = eta_adj x (-ln(0.70))^(1/beta_adj) - t_current`	Remaining useful life (Weibull P30)	Module D

Severity Level Boundaries

Score Range	Level	Meaning
0.00-0.30	normal	No significant fault indication
0.30-0.50	watch	Early indication — monitor closely
0.50-0.80	warning	Clear fault signature — plan corrective action
0.80-1.00	alarm	Critical condition — act immediately

The domain frameworks, factor catalog, and rule reference documented in this chapter provide the complete causal intelligence that separates RAPID AI from pattern-matching systems. The next chapter shows how this intelligence is organized into a dynamic RCM decision framework that continuously adjusts maintenance strategy based on real-time sensor data.

Next: Chapter 9 — RCM Framework Previous: Chapter 7 — The IMS