EVIDENCE. METHODS. TRACEABLE INTERPRETATION.
Research Methods for Monitoring Intelligence
RAUZ applies transparent research methods to monitoring QA/QC, cross-source validation, InSAR analysis, uncertainty assessment and engineer-reviewed interpretation across ground, infrastructure and environmental systems.
Research & Methods
A repeatable method for turning monitoring evidence into technical understanding.
RAUZ Research & Methods defines how monitoring questions are framed, how data quality is tested, how independent observations are compared and how uncertainty is kept visible before an engineering or environmental interpretation is reported. The aim is not to make every project use the same instrument or model. It is to make the reasoning behind the conclusion traceable.
Define what needs to be explained.
A useful method starts with a physical or operational question: movement, pore pressure, settlement, vibration, environmental change, trigger exceedance, sensor behaviour or another observable condition.
Separate observation from interpretation.
Raw readings, processed values, survey observations, InSAR products, environmental records and project events are treated as evidence with different resolutions, uncertainties and failure modes.
Keep the reasoning open to challenge.
Assumptions, exclusions, baselines, transformations, analytical settings and limitations should remain visible so another competent reviewer can understand how the conclusion was reached.
Method Framework
Question → evidence → validation → interpretation.
The RAUZ method is intentionally simple enough to audit and flexible enough to work across geotechnical, environmental and remote-sensing datasets. The exact calculation, instrument mix and review frequency change with the project; the logic below should not.
- Use the minimum evidence needed to answer the question well.
- Preserve raw data and identify every transformation.
- Do not interpret an unexplained spike before checking data quality.
- Use independent evidence where the consequences justify it.
- Review trend and rate, not only absolute magnitude.
- State uncertainty instead of filling missing context with assumptions.
Evidence Architecture
Choose evidence for the mechanism—not for the dashboard.
Instrument selection should follow the monitoring objective, anticipated mechanism, required spatial and temporal resolution, site access and project response needs. The table below illustrates evidence combinations; it is not a project design specification.
| Question | Possible primary evidence | Useful independent context | Method question |
|---|---|---|---|
| Lateral ground or wall movement | Inclinometer, survey prism / total station, GNSS where suitable | Excavation stage, strut level, groundwater, InSAR for wider-area context | Does the shape, direction, rate and neighbouring evidence support a credible movement mechanism? |
| Pore pressure / groundwater response | Vibrating-wire or standpipe piezometer, water-level records | Rainfall, pumping / dewatering logs, tidal or river influence where relevant | Is the pressure response physically plausible and temporally consistent with the hydraulic context? |
| Settlement or heave | Levelling, settlement points, hydrostatic settlement system, GNSS where suitable | Construction loading, groundwater, InSAR, adjacent structures | Is the movement local, regional, progressive, seasonal or linked to a project event? |
| Rotation, cracking or structural response | Tiltmeter, crackmeter, strain / load measurement, survey | Temperature, vibration, operational changes, support conditions | Is the response stable, cyclic, event-driven or part of a wider deformation pattern? |
| Environmental condition | Weather, water level / quality, soil moisture, noise, vibration or other scope-specific sensors | Site setting, land use, rainfall, asset operation, sampling / maintenance record | Is the observed change real, representative and relevant to the environmental question being asked? |
Data QA/QC
Before asking what a trend means, test whether the trend can be trusted.
ISO 18674-1 includes data processing, verification and reporting within the general monitoring framework. WMO likewise treats availability, timeliness, metadata and quality as distinct parts of observation quality. RAUZ applies the same discipline to monitoring intelligence: data quality is part of the technical method, not a final spreadsheet check.
Instrument & metadata
Instrument ID, location, depth, orientation, units, reference system, sensor / logger identity and installation history.
Reference control
Approved baseline, re-zero events, replacement history, survey control, calibration / function checks and known changes in configuration.
Completeness & validity
Missing periods, duplicates, impossible values, flat-lines, timestamp anomalies, unit mismatch, spikes, steps, drift and communication outages.
Processing record
Raw values, corrections, filtering, exclusions, calculation windows, thresholds, version history and reviewer decisions kept distinguishable.
Cross-Source Validation
Confidence rises when independent evidence tells a compatible story.
Cross-validation does not mean forcing two different measurement systems to agree. It means understanding what each system observes, in which coordinate or line-of-sight direction, at what spatial and temporal resolution, and whether the combined evidence supports or challenges the same physical interpretation.
Instrument ↔ survey
Compare profile or point movement with independent geodetic observations, while checking reference stability and whether the two systems observe the same component of movement.
Pore pressure ↔ rainfall / dewatering
Compare groundwater response with rainfall, pumping, excavation or hydraulic events without assuming causation from timing alone.
InSAR ↔ ground observations
Use satellite-derived motion to add spatial and historical context, then compare with GNSS, levelling, survey or instrumentation where geometry and coverage permit.
Movement ↔ project chronology
Overlay construction stages, loads, excavation, support installation, maintenance, weather or operational events to test plausible mechanisms and identify competing explanations.
Monitoring Analytics
Use analytics to organise evidence, not to hide engineering judgement.
RAUZ uses analytical methods according to the data quality, sampling interval, monitoring objective and physical question. A statistically unusual point is not automatically an engineering event, and a smooth trend is not automatically reliable.
Trend, rate and acceleration
Review magnitude, direction, persistence, rate of change and, where justified, acceleration. Window length and calculation method should be stated rather than hidden.
Steps, shifts and change points
Screen for sudden changes, new regimes, baseline discontinuities or behaviour that begins after a known project or environmental event.
Cross-sensor comparison
Compare related sensors, nearby locations, survey observations, groundwater, weather or project events where the relationship is physically meaningful.
Outlier investigation
Use range, rate, persistence, neighbouring evidence, instrument history and field context to distinguish a suspicious value from a credible change.
Recurring environmental patterns
Where sufficient history exists, separate daily, seasonal or weather-driven variation from structural or ground behaviour that warrants a different response.
Screening and report support
AI can assist triage, comparison and repetitive reporting, but the inputs, limitations and consequential interpretation remain subject to engineering review.
InSAR & Remote Sensing
Treat satellite ground motion as a measurement geometry—not a universal truth layer.
NASA/JPL explains that InSAR derives surface motion from phase differences between repeated radar observations, with displacement observed along the radar line of sight. Copernicus EGMS uses Sentinel-1 InSAR to provide consistent ground-motion information across Europe. RAUZ uses such products as complementary evidence where the site, geometry, coherence, temporal coverage and engineering question make the method suitable.
Screen the product
Review source, reference frame, acquisition period, line-of-sight geometry, processing method, spatial density, coherence / quality indicators and known gaps.
Match the scale
Wide-area satellite movement may describe a different spatial phenomenon from a local prism, inclinometer or crackmeter. Comparison must respect scale and direction.
Ground-check the interpretation
Where available, compare satellite patterns with survey, GNSS, levelling, instrumentation, groundwater, geology and project events before assigning a mechanism.
Project Context
The same method changes when the geology, climate, asset and failure mechanism change.
RAUZ operates across the South Caucasus, European Union, United Kingdom, Middle East and Africa, but a global method cannot substitute for local facts. Project-specific work should use official or client-controlled sources for site geology, stratigraphy, groundwater, climate, construction sequence, environmental setting, statutory requirements and asset history.
Geology & stratigraphy
Soil and rock sequence, weak layers, weathering, fill, faults, karst or other features can change the expected deformation mechanism and the depth or location of useful monitoring.
Groundwater & hydrology
Piezometric level, seepage, dewatering, rainfall, river or tidal influence can affect both ground behaviour and how a change should be interpreted.
Weather & environmental cycles
Temperature, rainfall, drought, freeze-thaw, humidity or seasonal effects can influence structures, ground and sensors. Relevant variables should be recorded rather than treated as background noise.
Project type & failure mode
A tunnel, excavation, slope, dam, bridge, mine, building, agricultural system or environmental site requires a different monitoring question and evidence hierarchy.
Construction or operational chronology
Excavation stages, support installation, loading, pumping, blasting, maintenance and operating changes can provide the event timeline needed to test a mechanism.
Standards, triggers & responsibility
Trigger levels, reporting frequency, response times, professional sign-off and decision authority must come from the governing project documents and appointments.
Traceability & Reproducibility
A useful monitoring conclusion should be reconstructable.
Monitoring intelligence becomes difficult to defend when raw data, corrections, thresholds and review decisions are mixed together. RAUZ therefore treats provenance and revision history as part of the method.
- Retain the original source file or immutable raw-data reference.
- Record units, coordinate system, datum, depth reference and time zone.
- Identify calibration, maintenance, replacement and re-baselining events.
- Document cleaning, filtering, interpolation and excluded records.
- State rate-calculation windows and analytical parameters.
- Separate project-approved trigger criteria from analytical screening rules.
- Preserve reviewer comments, status and revision history.
- Distinguish measured fact, derived value, interpretation and recommendation.
Official Technical Context
Primary sources that inform the method.
The references below are official public sources used as technical context. They do not imply certification, endorsement, partnership or that one standard governs every RAUZ engagement.
General performance-monitoring rules covering the ground, ground-interacting structures and geotechnical works, including requirements for measurements, data processing, verification and reporting.
Instrument-specific standard for displacement measurements across a line using inclinometers in geotechnical monitoring.
Open, geospatial-enabled standard for connecting IoT devices, observations and metadata from heterogeneous sensor systems.
European ground-motion information derived from Sentinel-1 InSAR, with product documentation and datasets for natural and anthropogenic ground movement.
Official explanation of how repeat-pass SAR interferometry uses phase differences to reveal surface motion and deformation.
Standards, procedures and quality-control guidance for environmental and meteorological observing methods across varying infrastructures and climatic conditions.
Official USACE guidance for instrumentation, monitoring and assessment of embankment dam and levee performance.
Research Outputs & Collaboration
Turn methods into reviewable technical assets.
RAUZ research work is intended to improve repeatability, transparency and technical usefulness in live monitoring workflows. Depending on the engagement, the output can be a project review, analytical method, technical demonstration, workflow specification or a jointly developed data / monitoring method.
Documented analytical method
Question, inputs, QA/QC, calculation logic, assumptions, limitations, validation evidence and expected output in a reviewable format.
Reproducible demonstration
Use public, anonymised or synthetic data—clearly labelled—to demonstrate anomaly investigation, cross-source comparison, reporting or another method.
Applied monitoring intelligence
Apply an agreed method to client monitoring data, project context and specific technical questions under a defined scope and review responsibility.
Potential collaboration — instrumentation & data providers
Potential collaboration — InSAR & remote sensing
Potential collaboration — universities & research groups
Potential collaboration — local engineering partners
Frequently Asked Questions
Research & methods questions.
Does RAUZ use one fixed analytical method for every project?
How does RAUZ decide whether an abnormal reading is real?
Can RAUZ combine InSAR with ground instrumentation?
Does AI make the final engineering decision?
Can RAUZ research methods be applied to environmental monitoring?
Can a client ask RAUZ to develop a project-specific method?
Technical Discussion
Bring the difficult monitoring question—not just the dataset.
A useful first discussion explains what changed, why it matters, what evidence is available and what decision the project team is trying to make. RAUZ can then define whether the next step is data diagnostics, independent review, InSAR interpretation, a monitoring-method study, recurring monitoring intelligence or another specialist input.
Useful starting material
Project brief, monitoring plan, instrument register, sample data, trigger table, relevant drawings, chronology, official geology / groundwater / environmental information and any existing interpretation that needs review.
Research question
State the question as clearly as possible: Is the movement credible? Why are two datasets conflicting? Is the rate increasing? Does the InSAR pattern match ground evidence? Is the monitoring method sufficient for the decision?