AUTOMATIC DETECTION

Trained on Massive Amounts of Data

Our models are capable of distinguishing more than 300 main- and subclasses of defects, which enables the reporting of all damages required by the European, British, and US coding standards. 

Our neural networks are trained on massive amounts of data: millions of images gathered in hundreds of different cities around the world.

Learn more about the Technology behind Pallon

WORKFLOW

Smoothly integrated into your Workflow

Just inspect the sewers the same way as always, except without spending time on coding the defects. Instead upload your inspection footage and the corresponding metadata to our webapp, where we automatically report all defects and deliver you the results within days.

Detection

A customized deep neural network, which has been trained on millions of data points, detects defects according to the coding standard (e.g. fissure, complex).

Quantification

By combining the output of neural networks and 3D vision we quantify the defects whenever applicable: Height and width, intrusion length, cross-section reduction as well as diameter reduction.

Localization

Using 3D vision we locate the defect’s longitudinal position (e.g. 9.1 meters) and circumferential position (e.g. 2-3 o’clock). If relevant, additional information such as continuous defect, at pipe joint, at connection and pipe wall are reported.

WHAT HAPPENS BEHIND THE SCENES

1

Simply keep inspecting sewers as you’ve always been doing, just without coding the defects.

Inspect

2

After inspecting a sewer, upload the inspection footage and the corresponding metadata to our webapp.

Upload

3

Once uploaded, we run our algorithms to detect, quantify and localize defects according the coding standard of your choice.

Analyze

4

You get notified when the report, condition grading and photos are ready, so you can share them using our web viewer or just download them.

Share