Case Studies

Every number comes from field verification and on-site measurement — trusted by leading energy and infrastructure operators.

Field Verified

PetroChina — high-sulfur gas pipeline

Third-party intrusion and leak early-warning on a high-sulfur pipeline — replaced a leading global vendor after a multi-vendor field trial.

100%
DAS recognition accuracy (field test)
Long-haul Pipeline

National operator — China-Myanmar gas line

59.14 km gas spur line; excavator 25 m, farming 5 m, manual 2 m alarm ranges verified.

95%
Effective alarm rate
CCUS

PetroChina — CCUS CO₂ pipeline

First CCUS CO₂ long-haul pipeline third-party intrusion and leak early-warning system.

1st
CCUS CO₂ pipeline application
First-of-kind

Sinopec — high-sulfur pipe-in-pipe

First high-sulfur pipe-in-pipe third-party intrusion and leak early-warning system.

1st
High-sulfur pipe-in-pipe application
Procurement

PetroChina — pipeline fiber sensing

Ranked first in centralized procurement; 32 of 33 alarms verified valid.

≥96%
Effective alarm rate
Infrastructure

Shudao Group — highway tunnel

Structural health monitoring for the largest cross-section highway tunnel.

Largest
Cross-section tunnel monitored
How We Measure

How our numbers are measured

What do the figures on this page actually mean? "Recognition accuracy" refers to DAS event identification during on-site field testing, where every alarm produced by the system is checked against what really happened on the ground. "Effective alarm rate" is measured over long-term operation: of all alarms raised, the share confirmed as genuine events rather than nuisance triggers. Because different activity types justify different response urgency, alarm zones are calibrated per event class — on the China–Myanmar gas spur line described below, the verified ranges were 25 m for excavator operations, 5 m for farming activity and 2 m for manual activity. Every number published here comes from field verification and on-site measurement on the customer's asset, not from laboratory simulations.

Case 01

PetroChina — high-sulfur gas pipeline

Why does a high-sulfur gas pipeline demand stricter intrusion and leak detection than an ordinary one? Hydrogen-sulfide-bearing service is corrosive, hazardous to people and environmentally sensitive, so any third-party activity near the right-of-way must be identified early and located precisely. On this PetroChina pipeline, Landsub Global's DAS system was evaluated in a multi-vendor field trial alongside a leading global vendor — and Landsub Global's system achieved 100% on-site event recognition during the field test, after which it replaced the incumbent international vendor on the line. For the operator, the result demonstrated that the full in-house chain — optical path, demodulation and AI event classification — could outperform an established supplier under the most demanding service conditions, on the customer's own asset and on the customer's own test protocol.

Case 02

National operator — China–Myanmar gas pipeline spur

Can a fiber optic sensing system stay reliable across a long, rural right-of-way with constant agricultural and construction activity? On a 59.14 km gas pipeline spur forming part of the China–Myanmar corridor, operated by the national pipeline operator, Landsub Global's DAS system was deployed for continuous third-party intrusion and leak early-warning. Over long-term operation the system sustained a 95% effective alarm rate — the share of alarms confirmed as genuine events — across a right-of-way that mixes farmland, roads and populated areas. Alarm zones were calibrated per activity class and verified on site: 25 m for excavator operations, 5 m for farming and 2 m for manual activity. This is the operating profile where nuisance alarms usually defeat competing systems; sustained 95% effectiveness across 59.14 km of mixed terrain is what field-tuned AI classification delivers.

Case 03

PetroChina — CCUS CO₂ pipeline

What does it take to monitor a CO₂ pipeline for carbon capture, utilization and storage? CCUS CO₂ pipelines are a young asset class: the transported fluid is dense-phase, odorless and harder to trace than natural gas, so third-party intrusion and leak early-warning has to come from the sensing system itself. Landsub Global delivered the intrusion and leak early-warning system for PetroChina's CCUS CO₂ long-haul pipeline — the first application of third-party intrusion and leak early-warning on a CCUS CO₂ long-haul pipeline in China. The deployment extended distributed fiber optic sensing from conventional hydrocarbon service into carbon-management infrastructure, and it was followed by further first-of-kind applications, including high-sulfur pipe-in-pipe monitoring for Sinopec described below.

References

Other selected references

  • Sinopec — high-sulfur pipe-in-pipe: the first third-party intrusion and leak early-warning system applied to a high-sulfur pipe-in-pipe pipeline.
  • PetroChina — centralized procurement: during procurement acceptance, 32 of 33 alarms were verified as valid — an effective alarm rate of ≥96%.
  • Shudao Group — highway tunnel: structural health monitoring for the largest cross-section highway tunnel.
Takeaways

What these projects have in common

Across these references, three practices repeat on every project:

  1. Field calibration before go-live. Alarm thresholds and AI event models are tuned against real signatures on the customer's terrain — soil, traffic, farming, construction — before the system takes operational duty.
  2. Verification against ground truth. Every acceptance figure in this page is the product of alarm-by-alarm checking against actual events, recorded by the customer's own test or acceptance protocol.
  3. Long-term operation, not one-off demos. The 95% and ≥96% effective alarm rates were sustained over continuous service on live pipelines, which is a harder test than any trial.
Transferability

Where the same evidence transfers

Do these results only apply to oil and gas pipelines? No — the underlying platform is the same. The DAS units that delivered 100% field-test recognition on a high-sulfur pipeline and 95% effective alarm rate across 59.14 km are the same hardware and AI stack that Landsub Global deploys on power cable tunnels, railway and highway structures, conveyor systems and live optical cable networks. What changes between sectors is the event model and calibration, not the sensing chain. For example, the largest cross-section highway tunnel we monitor uses our structural monitoring approach on the DTSS-BOTDR side, while in-service fiber monitoring keeps telecom and composite cables under 7×24 watch to 100 km with ±1 m fault location. When you evaluate us for a new sector, the pipeline references on this page serve as engineering evidence of range, sensitivity and alarm discipline — and our job in the proposal is to show how the same field-calibration method maps onto your asset.

Comparison

Why operators switched from conventional systems

A factual comparison against conventional point-based systems — the incumbents named here stay anonymous; the operating characteristics do not.

Field evidence: conventional point-based systems vs distributed fiber monitoring
DimensionConventional point-based systemsDistributed fiber monitoring on these projects
Route coverageInstrumented points with long unmonitored gaps along the right-of-wayContinuous coverage of every meter — 59.14 km supervised on a single gas spur line
Alarm credibilityNuisance alarms typically erode operator trust in the system95% effective alarm rate sustained over long-term operation; 100% event recognition in a multi-vendor field trial
Event localizationAccuracy degrades with distance from the nearest instrumented pointMeter-level localization along the full route, verified per activity class on site
Trial outcomeIncumbent systems evaluated on identical test protocolsSelected to replace a leading global vendor after a customer-run multi-vendor field trial
Verification

How we verify every case

Every figure published on this page passes the same acceptance discipline — here is what that looks like on site.

  1. Joint test plan with the operator. Before go-live, the customer defines the event classes, alarm zones and acceptance protocol on their own asset — we are measured by their standard, not ours.
  2. Ground-truth excitation. Real activities — excavation, farming, manual digging, vehicle movement — are performed along the route while the system reports, so every alarm can be checked against what actually happened.
  3. Alarm-by-alarm review. Each alarm raised during the acceptance window is classified as valid or nuisance and recorded, which is how figures such as 32 of 33 alarms verified valid are produced.
  4. Long-term operation tracking. After acceptance, the effective alarm rate is tracked in continuous service — the 95% figure on the China–Myanmar gas spur line comes from live operation, not from the trial window.
  5. Documented handover. Results are written into the customer's own acceptance records, so every number we quote on this page can be traced back to a documented field measurement.
FAQ

Frequently asked questions

Why was the incumbent vendor replaced on the high-sulfur pipeline?

During a multi-vendor field trial run by the operator on the same line, Landsub Global's DAS system achieved 100% on-site event recognition under the customer's own test protocol — after which it replaced a leading global vendor. The comparison was conducted on identical fiber, terrain and event classes, so the deciding factor was measured alarm quality, not specification sheets.

Do these pipeline results transfer to other assets?

The sensing platform is the same across sectors; what changes is the event model and site calibration. The same DAS hardware and AI stack that delivered 95% effective alarm rate across 59.14 km also runs on power cable corridors, railway structures, conveyors and live optical cable networks — with RFTS keeping in-service fibers under 7×24 watch to 100 km with ±1 m fault location.

Want to see how it applies to your asset?

Tell us your scenario — we'll share relevant references.

Talk to Us

Last updated: September 2026