top of page
Search

Beyond the Data: Turning Geospatial Information into Actionable Insight

Writer: Jesskiran Bhatia
Jesskiran Bhatia
Nov 10
3 min read

The data never really stops. Earth observation satellites pass over the same place every one to six days, at least a dozen GNSS satellites are overhead at any moment, and on the ground, sensors track their own positions around the clock. We have never had more geospatial data, so the challenge is no longer collecting it. It is knowing what the signals are telling us, and when they matter.



More Data Doesn't Automatically Mean Better Insight


Each geospatial technology sees the world in a different way, and each has blind spots.


Interferometric Synthetic Aperture Radar (InSAR) compares radar images of the same area taken days or weeks apart. Tiny differences in the returning signal reveal ground movement of just a few millimetres, across areas far larger than any survey team could cover. But InSAR only measures movement along the satellite's line of sight, a slanted view rather than straight up and down. It also works best on stable, hard surfaces such as buildings and roads; over fields, woodland or snow, the signal can become too noisy to use.


GNSS (the family of satellite positioning systems that includes GPS) fills some of those gaps. A fixed GNSS station records its exact position in three dimensions, continuously, but only at that one point. Optical satellite imagery adds visual context, such as new construction, flooding or changes in vegetation, but it cannot detect millimetre-scale movement and, in the UK, is frequently blocked by cloud.


None of these tells the full story on its own. A single InSAR measurement might show a building moving, but without context it is hard to say whether that movement matters.



Making Datasets Work Together


Combining these sources is where the real value lies, but it is rarely straightforward. Even when datasets cover the same site, they differ in how often they are collected, how much detail they capture, how accurate they are, and the coordinate system they use. Interoperability, the ability to exchange and combine data from different systems, means resolving those differences and understanding each measurement's uncertainty before comparing them.


Get this right and the data starts to answer real questions. On clay soils, common across much of England, the ground naturally swells in wet winters and shrinks in dry summers, rising and falling by several millimetres each year. Compare a single measurement from February with one from August, and a perfectly stable site could look like it is sinking. Only by combining a longer record of satellite measurements with ground-based positioning and knowledge of the local geology can normal seasonal behaviour be separated from genuine, ongoing subsidence.



Project CORE: Bringing the Pieces Together


GVL's Project CORE, supported by the European Space Agency's InCubed programme, is designed to do exactly this. It combines InSAR, optical imagery and GNSS-related measurements into a single, more reliable picture of how the ground and the structures on it are moving. This is especially valuable for buildings, bridges and transport networks, where slow movement can be an early sign of structural or ground stability problems, and where asset owners currently rely on periodic site inspections that are costly and hard to carry out at scale.


Artificial intelligence and machine learning (AI/ML) sit at the heart of CORE's approach, known as All-Terrain InSAR (AT-InSAR). As described above, conventional InSAR struggles over fields, woodland and other vegetated ground, leaving gaps in coverage exactly where many roads, railways and pipelines run. CORE uses AI/ML to fill those gaps, interpolating reliable movement estimates in areas where the radar signal alone is too noisy, and producing far more complete coverage than InSAR on its own.


The same techniques help make sense of the volume of data involved. A single site can contain thousands of measurement points, each with years of history. Machine learning can group these points by how they behave over time, for example separating points that move up and down with the seasons from those showing a steady downward trend, and flag the unusual ones for a closer look.


The aim is not to replace engineers or detailed investigation, but to point them to the right places. Rather than simply detecting movement, CORE aims to help organisations understand whether it matters, prioritise inspections and plan maintenance with confidence.



Making Geospatial Data Work Harder


Accurate measurement will always be the foundation of geospatial work. But as data becomes more plentiful, the advantage shifts to those who understand its limits, connect the right sources and present the results clearly. The future of geospatial technology is not about collecting more information; it is about making better use of what we already have.

 
 
 

Comments


image.png

Offices

Geospatial Ventures Ltd 

Sir Colin Campbell Building 

Jubilee Campus Nottingham 

NG7 2TU

Charlotte House

Stanier Way 

The Wyvern Business Park

Derby

DE21 6BF

Contact

operations@geospatialventures.co.uk

​

​

+44 (0) ‭7973 139368‬

​

​

07881 336841​​

Follow

  • 1_edited
  • 2_edited
  • 3_edited
  • 4_edited

@geospatialventureslimited

​​​

​

@GVL_Solutions

​​

​

​​@geospatialventures​​

​

​

Geospatial Ventures Blog

bottom of page