


Acoustic edge leak detection for unmetered pipelines.
Acoustic edge leak detection for unmetered pipelines.
Acoustic edge leak detection for unmetered pipelines.
DPx combines high-rate pressure sensing, spectral analysis, and on-site machine learning to identify leak signatures where traditional systems can’t.
DPx combines high-rate pressure sensing, spectral analysis, and on-site machine learning to identify leak signatures where traditional systems can’t.
DPx combines high-rate pressure sensing, spectral analysis, and on-site machine learning to identify leak signatures where traditional systems can’t.



0
0
1
2
3
4
5
Here’s how DPx works.
Here’s how DPx works.
Here’s how DPx works.
DPx detects leaks by analyzing high-frequency pressure signals directly at the pipeline, using on-site machine learning to isolate and classify acoustic leak signatures in real time.
DPx detects leaks by analyzing high-frequency pressure signals directly at the pipeline, using on-site machine learning to isolate and classify acoustic leak signatures in real time.
DPx detects leaks by analyzing high-frequency pressure signals directly at the pipeline, using on-site machine learning to isolate and classify acoustic leak signatures in real time.

High-rate dynamic pressure sensing
3-4 weeks
Dynamic pressure sensors capture high-frequency acoustic data that propagate along the pipeline during abnormal events
Enables sensitivity to leak-generated signatures outside the bandwidth of standard SCADA data
High-rate dynamic pressure sensing

High-rate dynamic pressure sensing
3-4 weeks
Dynamic pressure sensors capture high-frequency acoustic data that propagate along the pipeline during abnormal events
Enables sensitivity to leak-generated signatures outside the bandwidth of standard SCADA data
High-rate dynamic pressure sensing

High-rate dynamic pressure sensing
3-4 weeks
Dynamic pressure sensors capture high-frequency acoustic data that propagate along the pipeline during abnormal events
Enables sensitivity to leak-generated signatures outside the bandwidth of standard SCADA data
High-rate dynamic pressure sensing

Spectral and statistical signal processing
3-4 weeks
Raw acoustic waveforms are transformed into spectral features that help isolate leak-related patterns from pump noise, vibration, and operational variability.
Enables clean feature inputs for the ML model and robust performance in noisy environments.
Spectral and statistical signal processing

Spectral and statistical signal processing
3-4 weeks
Raw acoustic waveforms are transformed into spectral features that help isolate leak-related patterns from pump noise, vibration, and operational variability.
Enables clean feature inputs for the ML model and robust performance in noisy environments.
Spectral and statistical signal processing

Spectral and statistical signal processing
3-4 weeks
Raw acoustic waveforms are transformed into spectral features that help isolate leak-related patterns from pump noise, vibration, and operational variability.
Enables clean feature inputs for the ML model and robust performance in noisy environments.
Spectral and statistical signal processing

On-site machine learning classification
3-4 weeks
A lightweight ML model evaluates each set of extracted features and assigns leak / non-leak labels in real time.
Enables autonomous detection without the bandwidth, latency, or reliability constraints of remote/cloud processing.
On-site machine learning classification

On-site machine learning classification
3-4 weeks
A lightweight ML model evaluates each set of extracted features and assigns leak / non-leak labels in real time.
Enables autonomous detection without the bandwidth, latency, or reliability constraints of remote/cloud processing.
On-site machine learning classification

On-site machine learning classification
3-4 weeks
A lightweight ML model evaluates each set of extracted features and assigns leak / non-leak labels in real time.
Enables autonomous detection without the bandwidth, latency, or reliability constraints of remote/cloud processing.
On-site machine learning classification

Rolling aggregation for decision confidence
3-4 weeks
Individual classifications are combined within a rolling time window to ensure alerts represent persistent and meaningful patterns, not single-frame transients.
Enables strong resistance to false alarms and stable decision-making.
Rolling aggregation for decision confidence

Rolling aggregation for decision confidence
3-4 weeks
Individual classifications are combined within a rolling time window to ensure alerts represent persistent and meaningful patterns, not single-frame transients.
Enables strong resistance to false alarms and stable decision-making.
Rolling aggregation for decision confidence

Rolling aggregation for decision confidence
3-4 weeks
Individual classifications are combined within a rolling time window to ensure alerts represent persistent and meaningful patterns, not single-frame transients.
Enables strong resistance to false alarms and stable decision-making.
Rolling aggregation for decision confidence

Alerting and event notification
3-4 weeks
When a leak pattern is confirmed, DPx transmits an event through the operator’s preferred interface, typically SCADA, a PLC, or a secure browser view.
Only the final decision is sent, keeping the system lightweight and bandwidth-efficient.
Alerting and event notification

Alerting and event notification
3-4 weeks
When a leak pattern is confirmed, DPx transmits an event through the operator’s preferred interface, typically SCADA, a PLC, or a secure browser view.
Only the final decision is sent, keeping the system lightweight and bandwidth-efficient.
Alerting and event notification

Alerting and event notification
3-4 weeks
When a leak pattern is confirmed, DPx transmits an event through the operator’s preferred interface, typically SCADA, a PLC, or a secure browser view.
Only the final decision is sent, keeping the system lightweight and bandwidth-efficient.
Alerting and event notification
Optimal applications for DPx.
Optimal applications for DPx.
Optimal applications for DPx.
DPX is designed for liquid pipeline segments where flow-based or CPM systems cannot operate reliably due to limited instrumentation, short length, or high operational noise.
DPX is designed for liquid pipeline segments where flow-based or CPM systems cannot operate reliably due to limited instrumentation, short length, or high operational noise.
DPX is designed for liquid pipeline segments where flow-based or CPM systems cannot operate reliably due to limited instrumentation, short length, or high operational noise.
Short, unmetered pipeline segments
Laterals, stub lines, and small transfer lines without reliable flow measurement
3-4 weeks
Short, unmetered pipeline segments
01
Short, unmetered pipeline segments
Laterals, stub lines, and small transfer lines without reliable flow measurement
3-4 weeks
Short, unmetered pipeline segments
01
Short, unmetered pipeline segments
Laterals, stub lines, and small transfer lines without reliable flow measurement
3-4 weeks
Short, unmetered pipeline segments
01
River crossings and environmentally sensitive areas
Buried, inaccessible, or high-consequence segments where added instrumentation is impractical
3-4 weeks
River crossings and environmentally sensitive areas
02
River crossings and environmentally sensitive areas
Buried, inaccessible, or high-consequence segments where added instrumentation is impractical
3-4 weeks
River crossings and environmentally sensitive areas
02
River crossings and environmentally sensitive areas
Buried, inaccessible, or high-consequence segments where added instrumentation is impractical
3-4 weeks
River crossings and environmentally sensitive areas
02
Single-ended custody transfer and LACT lines
Short custody transfer segments where operators can read and control only one end of the line
3-4 weeks
Single-ended custody transfer and LACT lines
03
Single-ended custody transfer and LACT lines
Short custody transfer segments where operators can read and control only one end of the line
3-4 weeks
Single-ended custody transfer and LACT lines
03
Single-ended custody transfer and LACT lines
Short custody transfer segments where operators can read and control only one end of the line
3-4 weeks
Single-ended custody transfer and LACT lines
03
Remote or off-grid locations
Assets without reliable power, communications, or centralized control systems
3-4 weeks
Remote or off-grid locations
04
Remote or off-grid locations
Assets without reliable power, communications, or centralized control systems
3-4 weeks
Remote or off-grid locations
04
Remote or off-grid locations
Assets without reliable power, communications, or centralized control systems
3-4 weeks
Remote or off-grid locations
04
Want to explore how DPx could fit into your pipeline infrastructure?
Book a discovery call to walk through deployment scenarios, integration options, and field performance considerations.

Want to explore how DPx could fit into your pipeline infrastructure?
Book a discovery call to walk through deployment scenarios, integration options, and field performance considerations.

Want to explore how DPx could fit into your pipeline infrastructure?
Book a discovery call to walk through deployment scenarios, integration options, and field performance considerations.
