Where equipment behavior matters before failure
Each application maps a critical asset to its dominant failure mode, the gap in current monitoring, and how physics-based edge monitoring can help.
Commercial Marine Propulsion & Auxiliary Systems
Merchant vessels operate with limited maintenance windows. Main propulsion bearings and auxiliary turbochargers are critical assets where in-voyage failure can force emergency drydocking at enormous cost.
CRITICAL ASSET
Main propulsion shaft bearings, auxiliary turbocharger assemblies
FAILURE MODE
Bearing degradation, blade fatigue, lubrication breakdown under sustained load
MONITORING GAP
Shore-based condition monitoring depends on connectivity and introduces latency. Scheduled inspection may miss developing faults between port calls.
PRED-D SIGNAL
Continuous on-board edge analysis of vibration and acoustic signatures. Autonomous detection without shore connectivity.
OPERATIONAL DECISION
Plan port-of-opportunity maintenance vs. risk emergency drydocking
EVIDENCE STATUS
Proposed pilot scenario — not yet field-validated on operational vessels
High-Velocity Slurry Pipeline Transport
Mineral processing and mining operations transport abrasive slurry at high velocity. Pipe wall erosion progresses silently until critical wall thinning leads to rupture, environmental release, and production shutdown.
CRITICAL ASSET
Slurry pipeline segments, elbows, and reducers in high-velocity mineral transport
FAILURE MODE
Hydro-erosion wall thinning, impact wear at bends, internal coating degradation
MONITORING GAP
Periodic ultrasonic thickness testing is labor-intensive, provides only snapshots, and may not capture accelerating erosion between inspections.
PRED-D SIGNAL
Continuous acoustic emission monitoring to detect erosion progression and internal impact events
OPERATIONAL DECISION
Schedule segment replacement during planned shutdowns rather than emergency pipeline isolation
EVIDENCE STATUS
Engineering analysis — field validation planned for representative pipeline installations
Supercritical Boiler Feedwater Pumps
Feedwater pumps in supercritical thermal power plants operate under extreme temperature and pressure. Seal and bearing degradation can cascade into forced outage, costing millions in lost generation and emergency repair.
CRITICAL ASSET
Boiler feedwater pump bearings, mechanical seals, and drive couplings
FAILURE MODE
Seal degradation, bearing wear, cavitation damage under supercritical operating conditions
MONITORING GAP
Conventional vibration monitoring may not distinguish early-stage degradation from normal operating variation at supercritical parameters.
PRED-D SIGNAL
Physics-informed analysis tuned to supercritical operating regimes, distinguishing genuine degradation from process-induced signal variation
OPERATIONAL DECISION
Coordinate pump overhaul with planned outage windows rather than forced mid-cycle shutdown
EVIDENCE STATUS
Engineering analysis based on published rotary asset dynamics research. Field deployment not yet initiated.
Deployment considerations across sectors
Physical Access
Sensor mounting on operating equipment, access during shutdowns vs. runtime, environmental enclosure requirements.
Sensor Environment
Temperature, vibration, moisture, EMI, and chemical exposure considerations across marine, mining, and power environments.
Connectivity
Autonomous edge operation for connectivity-limited environments; periodic sync for fleet-level analytics.
Maintenance Integration
Output compatibility with existing CMMS, historian, and work-order systems.
What a pilot engagement looks like
Note: Proposed pilot scenario — this is not a completed deployment.
- Identify a critical asset and failure mode of interest.
- Install sensor nodes on or near the asset.
- Collect baseline data over 2–4 weeks of normal operation.
- Monitor for signal changes that correlate with developing conditions.
- Compare detection with existing inspection findings.
- Evaluate whether lead time, sensitivity, and false-alarm behavior meet operational requirements.
Is your asset a good fit?
- Asset criticality: Unplanned failure has significant safety, cost, or production consequences.
- Accessible signals: Vibration, acoustic, or thermal signals can be physically acquired from the asset.
- Known failure modes: The asset has documented degradation patterns that can be characterized.
- Maintenance process: A maintenance workflow exists that can act on early detection.
- Connectivity: Edge operation is available; periodic connectivity sufficient for reporting.
- Success criteria: Clear metrics for detection lead time, false-alarm rate, or maintenance cost reduction.
Application questions
Will it work with our existing instrumentation?
Our swarm nodes operate independently of your existing PLC/SCADA systems to ensure isolated, redundant data capture without interfering with control loops. While we don't rely on existing instrumentation for raw signals, our analysis can be correlated with your historian data if required.
How long does baseline collection take?
Typically 2 to 4 weeks of normal operational cycles, covering various load conditions. The exact duration depends on how frequently the equipment experiences its full range of operating states.
What does a pilot require from our side?
We require safe access during an installation window to mount the 30-pin ESP32 nodes, power provision (if not battery operated), and engineering time to define the operational thresholds, failure history, and maintenance workflow.
What if our failure mode isn't listed here?
We evaluate new applications continuously. If the failure mode creates a measurable change in physical behavior (vibration, acoustics, or temperature) that precedes failure, we can investigate its suitability for topological edge analysis.