• Careers
  • Fugro Global
  • Global
Innovation
07 April 2022 By Pooja Mahapatra Solution Owner Geospatial

The world's landscape is changing fast, with urban expansion and the increasing frequency of extreme weather events. To manage these changes effectively, government agencies and asset owners are increasingly demanding better geospatial data – because only by consulting accurate Geo-data can they prepare confidently for the future.  

Light Detection and Ranging (Lidar) is commonly used to map bare earth, hydrologic features, infrastructure and vegetation. In recent years the demand for high-quality lidar data at scale has increased.

Each iteration of lidar collection over an area of interest (AOI) provides insight from the acquired Geo-data. This helps us to characterise the earth at the time of capture, as well as to detect and predict change in any measured feature between two lidar acquisitions.

Automatic classification – an exciting development

Fugro’s advancement in lidar production processes make it possible to classify clusters of lidar points to identify unique earth features (such as bare earth, hydro, buildings, vegetation, culverts and utility assets) using machine-learning techniques at scale without compromising on speed, quality or cost.

Fugro’s Sense.Lidar® uses a combination of lidar point clouds, cloud processing, artificial intelligence (AI) and human expertise to improve schedules and reduce costs through production efficiencies.

Putting Sense.Lidar® to work

In March 2021 Fugro was commissioned to enhance the existing United States Geological Survey (USGS) lidar data from the federal specification to the state specifications of Texas for lidar classifications at a 99% accuracy. The survey area spanned 83,184 square miles (215,446 square kilometres).

We used the Sense.Lidar® machine-learning process to enhance the data from standard USGS classifications to add buildings, vegetation, and culverts – these key features enabled users to make sense of what they are viewing.

We classified the point cloud data 39% faster than would have been possible using traditional techniques, while also increasing quality and accuracy – with no compromise on cost.

Recurring lidar programmes add value

Many government agencies rely on lidar datasets to help them with their land-use planning and flood plain management activities. A single, well classified, lidar dataset can provide highly useful information about an area at the time of data capture. But it can’t be used to compare the present with the past.

Geo-data users are starting to recognise the economic value of increasing the frequency of their lidar programmes. They are requesting recurring programmes to help them detect human-made and environmental changes more quickly.

Closely recurring datasets can be vitally important in helping these users to understand and maintain assets, while also protecting the environment and safeguarding life.

The increased use of better lidar data to analyse the earth, measure and predict future changes requires significant financial investment. So, it is vital to extract maximum value from the multiple lidar datasets.

Traditional lidar data processing techniques (human classification or macros) are expensive, time-consuming and often inaccurate. Annually recurring lidar programs at scale require fast, accurate and automated classification technology.

Faster, better and more accurate

Having multiple classified lidar datasets that geo-locate and characterise assets and natural features improves the accuracy of Geo-data analysis.

Lidar-derived digital twins make the 3D Geographic Information System (GIS) more user-friendly, enabling better, more informed decisions to be made from the desktop. This approach:

  • Reduces operational costs – physical inspections typically account for around 20% of total survey project costs
  • Improves safety – fewer boots on the ground
  • Supports effective planning
  • Aids faster decision-making.

Efficiently and accurately assists with creating lidar-derived 3D digital twins of the human- and natural-built environments

Efficiently and accurately assists with creating lidar-derived 3D digital twins of the human- and natural-built environments

Sense.Lidar® can be deployed at any level – city, state or even country. Potential applications include flood risk management, forestry analysis, energy operations, infrastructure planning, agriculture monitoring, broadband management and various design programmes.

In conclusion

As time goes on, the speed, reliability and accuracy of automated lidar point classifications will only increase. And as the number of datasets increases, Geo-data users will be able to make better decisions and more effective preparations for what lies ahead.

Can be localised or scaled to city, county to country-wide analyses through cloud-based processing

Can be localised or scaled to city, county to country-wide analyses through cloud-based processing

Sense.Lidar - before and after

 Sense.Lidar® - before and after

Did you know?

  • Sense.Lidar® can be deployed at any scale – city, state or even country.
  • Sense.Lidar® point classifications were found to be 39% faster than traditional techniques

 

Most popular articles

Sign up for email updates