At Tiecken Solutions, we leverage advanced Computer Vision, 3D LiDAR analysis, and Document Intelligence to turn complex visual data, point clouds, and engineering documents into actionable intelligence.
Many asset owners hold surveys, inspection imagery and engineering documentation that were used once and then archived. That data typically holds far more information than it was captured for.
We analyse point clouds, imagery and documents automatically and deliver asset data you can act on: a position, a measured value and the source image for every object. Our methods were developed on rail projects and apply to other linear infrastructure and industrial sites.
We build software that turns inspection imagery, 3D point clouds and engineering documentation into asset data automatically. Asset owners and engineering firms use it to carry out inspections, asset registration and safety checks faster, with less work in the field.
High-speed 2D image recognition & asset inspection
3D point clouds, clearance profiling & Scan-to-BIM
Autonomous AI agents & automated data pipelines
Document AI & technical specification analysis
Tensioners and droppers measured automatically from mobile laser scanning data; insulated rail joints detected and located on inspection imagery. No possessions and no site visits. Click any visual to enlarge.
A single survey job over a multi-track corridor returned 301 catenary tensioners, located and measured, without a return visit to site. Each comes with its measured drop, its position and a QA image, and every tensioner is checked from several scanners and passes.
Droppers are measured from their clamp on the messenger to their clamp on the contact wire. The laser returns on a dropper stop short of both clamps, so a length read from the returns alone falls about a third short. We measure to where the dropper actually attaches. Tested against independently labelled scans the model had never seen.
4 Images
Automated detection and high-precision positioning of insulated rail joints (IRJs) using two complementary localization methods: GeoTIFF mapping with top-down linescan cameras and triangulation across forward-facing camera feeds.
Computer vision, LiDAR and inspection solutions, developed for and with partners in the rail sector.
"Tiecken Solutions is the partner with which Strukton Rail develops and implements groundbreaking, railway-related innovations in a pragmatic way. Innovations that are conceived and rolled out together with our operations."
Managing Director Strukton Rail Nederland
In collaboration with IGL (Masters in Measurement), we developed advanced Computer Vision algorithms for automated detection and high-accuracy geographic positioning of insulated rail joints (IRJs) across railway networks.
Geodesy & Rail Engineering Firm
AI applications from our projects in image recognition, 3D point cloud analysis and GIS localisation. Click any visual to enlarge.
2 Images
Development of a Computer Vision model that automatically detects overhead line (catenary) components from high-speed imagery. Significantly accelerates inspection routines and ensures consistent object classification.
2 Images
Automated vision-based system to verify the position of catenary tension counterweights safely from a distance, removing the need for inspectors to enter active track areas.
Image segmentation application that detects insulation layers within rail joints. Using calibrated pixel scaling, the system calculates the exact minimum thickness and identifies structural wear.
Contact wire height and stagger measured against the rails every 25 cm, over hundreds of metres at a time, with nobody picking points by hand. Height is taken perpendicular to the rail plane, so cant does not distort it. The accuracy is that of the survey the measurement comes from.
2 Images
3D point cloud segmentation model that classifies rails, overhead lines, and vegetation. Calculates millimeter-precise clearances between vegetation and critical rail assets to optimize trimming schedules.
2 Images
Infrastructure asset identification using both Bird's-Eye-View (BEV) projections and direct 3D spatial object recognition—ideal when 2D camera footage is occluded or height metadata is essential.
Combining 3D point clouds with high-precision GPS and IMU data (pitch, yaw, and roll). A virtual vehicle gauge clearance model traverses the point cloud to identify potential infrastructure collisions.
Our models were developed on rail data. A different asset or sector mainly calls for new training data; the processing chain stays the same.
Conductors, towers and vegetation clearance from LiDAR line-inspection data. This builds directly on our wire and mast analyses for the overhead line.
Road furniture, signage and clearance from mobile mapping data. GPR analysis for asphalt layers and buried objects is in development.
Scan-to-BIM: converting point clouds into IFC models (R&D). Document Intelligence: AI agents and RAG pipelines that search specifications, standards and reports, fully on-premise where required.
Automated analysis of your point clouds and imagery as part of what you deliver to your client, as in our project with IGL.
You supply a bounded set of inspection imagery or LiDAR data. No new capture is needed.
We show, on your own data, what is automatically detectable and measurable, confidence scores included.
Model and pipeline are fitted into your existing inspection chain and systems.
Pick one question about your assets that you suspect your existing survey data already answers. Within that scope, we will show you what can be taken out of it.
Personal commitment, deep AI expertise, and extensive domain experience in rail engineering and civil infrastructure.
Specialized in Computer Vision, Deep Learning, 3D LiDAR processing, and building custom AI software for railway inspections and smart asset management.
Connect with Sjoerd on LinkedIn
Extensive experience in rail engineering, measurement technology, innovation, and process integration in complex infrastructure projects.
Connect with Jacques on LinkedIn