Vibration Spectral Analysis
FFT-based frequency analysis and envelope demodulation to identify bearing defect frequencies, gear mesh harmonics, and impeller pass frequency anomalies.
FFT · Envelope demodulation
Rotating Equipment Condition Monitoring
Detect rotating equipment failures before unplanned shutdown.
Midstreamly reads your existing SCADA historian and vibration monitor data — no new hardware — and issues anomaly alerts hours before a breakdown.
No hardware procurement · No PLC changes · Typical time-to-alert: 5 weeks
VIBRATION — COMPRESSOR-04 · BEARING-DE
ANOMALY DETECTED
12.4 Hz
1× running
0.42 in/s
RMS velocity
3.2σ
deviation
WARNING
alert level
Connects to your existing systems — no new hardware required
AVEVA PI System
GE Digital Cimplicity
AVEVA Wonderware
Bently Nevada System 1
Emerson AMS Machinery Manager
SKF @ptitude
~73%
Reduction in unplanned downtime across monitored pump stations
6–18 hr
Average advance warning window before failure threshold
<0.3%
False-positive rate on anomaly alerts — we publish this metric
40+
Equipment models in the baseline library at system activation
How it works
From historian to alert in 5 weeks
No rip-and-replace. No new sensors. Midstreamly reads the data your systems already collect.
01
Connect
Read-only API connection to your SCADA historian or vibration monitor. Supports OPC-UA, OPC-DA, OSIsoft AF, and REST. No PLC changes, no firewall exceptions for write access.
02
Baseline
14–21 days of normal-operation data establishes equipment-specific thresholds calibrated to your actual running conditions — not generic reference curves.
03
Detect
ML models score each asset continuously. Anomalies flagged by multivariate deviation plus trajectory — reducing false positives from single-variable threshold crossings.
04
Alert
Configurable notifications to DCS operator console, reliability engineer inbox, or automatic CMMS work order creation. Three severity levels: Advisory, Warning, Critical.
Platform capabilities
Engineered for the failure modes that matter in midstream
Every capability was built for oil and gas rotating equipment — not retooled from manufacturing or HVAC analytics.
SCADA Historian Integration
Native connectors for OSIsoft PI Asset Framework, OPC-DA/UA polling, and direct REST API access. Handles tag aliasing, historian downtime, and data-quality flags.
OSIsoft AF · OPC-UA · OPC-DARUL Prediction
Remaining useful life estimates using LSTM neural networks for trajectory modeling and Weibull survival modeling for failure probability curves with uncertainty quantification.
LSTM · Weibull · Cox-PHAnomaly Scoring
Multivariate isolation forest detects deviation patterns across temperature, vibration, and process variables simultaneously. Overlaid with configurable ISO 20816 severity bands.
Isolation Forest · ISO 20816Failure Mode Library
40+ baseline signatures for bearing outer/inner race, seal face wear, impeller erosion, surge precursors, cavitation onset, and motor electrical faults in rotating equipment.
40+ fault signaturesIntegrity Alignment
Designed to support PHMSA pipeline safety integrity management program documentation. Aligns with API 670 machinery protection and ISO 20816 vibration severity band conventions.
PHMSA IMP · API 670 · ISO 20816
Equipment coverage
Three equipment categories. One monitoring platform.
Midstreamly ships with pre-built failure-mode signatures for the rotating equipment classes that drive unplanned downtime in midstream operations.
Centrifugal Compressors
Surge and stall precursors
Radial and thrust bearing wear
Dry gas seal degradation
Impeller fouling and erosion
Avg. 14 hr advance warning on surge events
Centrifugal Pumps
Cavitation onset detection
Rolling element bearing wear
Mechanical seal face degradation
Impeller wear and imbalance
Aligns with API 610 pump reliability metrics
Gas Turbines & Motors
Thermal excursion and hot-section wear
Shaft bearing imbalance and misalignment
Motor winding electrical faults
Coupling and gearbox anomalies
Shaft vibration cross-correlated with process data
Deployment outcomes
What early detection is worth in midstream
Synthetic outcome projections based on equipment failure data patterns from active pilot deployments. Operator identities are not disclosed.
$2.1M
Gas processing operator · Permian Basin
Avoided compressor train shutdown cost. Surge event detected 14 hours before critical alert threshold. Controlled maintenance window scheduled vs. emergency repair.
$840K
Pipeline pump station operator · Gulf Coast
Cavitation onset detected on a mainline pump 22 hours before seal failure. Replacement pre-ordered, downtime reduced from 6 days to 18 hours, throughput impact minimal.
73%
NGL fractionation facility · Mid-Continent
Reduction in unplanned rotating equipment downtime over a 9-month monitoring period. Primary driver: bearing wear detection on agitator motors and refrigeration compressors.
Your SCADA data is already generating the signals. Let's analyze them.
Book a technical session with a reliability engineer on our team — not a sales rep. We'll review your historian configuration and map your equipment population to our failure mode library.
Book a Technical Demo