TBC: Leveraging AI for advanced feature extraction

March 5, 2025

In Trimble^® Business Center (TBC) office software version 2024.10, feature extraction continues to be enhanced, leveraging advanced AI technologies for more comprehensive analysis capabilities that can be applied across a range of workflows.

With several powerful tools that cater to the diverse needs of users, from customizable feature extraction with the new generic CAD point extraction to more comprehensive pavement inspection, TBC helps enhance the precision and efficiency of urban planning, infrastructure development and road maintenance projects.

Customizable CAD point extraction tool

In addition to pre-trained classifiers and pre-built feature extraction tools, TBC now allows users to train custom AI models for point cloud classification, automating the extraction of any object of interest. This enables users to build their own Intellectual Property, distinguishing their offerings in the market and providing unique services to their customers.

Training custom AI point cloud classification models is now powered by customizable feature extraction. The new generic CAD point extraction tool can automatically extract the accurate location and create a CAD point for any object of interest in the point cloud based on the point cloud region. This advancement allows users to move from custom classification to custom feature extraction seamlessly.

New outputs for mobile mapping customers

The Trimble mobile mapping systems can now create perspective-free imagery of the pavement surface through the orthorectification of back-facing camera images. Additionally, in TBC, we adjust the radiometry of images and mosaic them to create a 2D visualization of the road surface. This not only provides a more convenient way for visual pavement inspection, but also enables accurate measurements of features, such as cracking.

Fully automated lane lines extraction for Trimble MX data

Trimble MX users can now extract lane lines automatically without user interaction. An AI model extracts vectors from imagery captured with a spherical camera. Multiple lane lines visible in the images will be extracted and optionally centered on the lane line marking based on the intensity values from the point cloud for those who need survey-grade accuracy.

Advanced pavement analysis

Trimble Business Center offers a comprehensive workflow for automated pavement inspection based on Trimble mobile mapping data. The functionality detects, classifies and evaluates the severity of pavement defects, calculates Pavement Condition Index score and International Roughness Index. The workflow enables safe and timely pavement inspection which helps decrease the cost of asset ownership and guarantees transportation user safety.

Detailed crack detection: Pavement analysis in TBC now includes the capability to evaluate the severity of cracking, in addition to detecting and classifying various pavement crack types. With Ortho Lane, images from the back-facing camera can be orthorectified, allowing for accurate measurements. Each detected crack is automatically measured and assigned a customizable severity level.

*Pavement orthomosaic generated using Otho Lane*

Road shoulder drop-off: TBC can now analyze shoulder drop-off and evaluate whether it exceeds acceptable values defined in standards used by public agencies. The severity level for shoulder drop-off is customizable and can be adapted to any standard. Assessing shoulder drop-off is critical for transportation user safety, as substantial drop-offs can become serious driving hazards.

The tools you need to succeed

TBC v2024.10 continues Trimble's commitment to innovation and excellence in geospatial technology. By incorporating advanced AI capabilities, this release provides users with more comprehensive analysis tools and workflows, enhancing the accuracy and efficiency of their projects. Whether you are working on urban planning, infrastructure development, or road maintenance, TBC v2024.10 feature extraction tools equip you with the tools you need to succeed.