A Digital Riser in the engineering of FTTx networks

‍

A pilot project, conducted by HFC Allinq department in collaboration with several suppliers, included YUNGO with the role of producing the digital engineering of an FTTx network. Let us tell you this story.

‍

The purpose was to define and design the network connection of a business park in the Netherlands, in both traditional and digital engineering ways. At the same time, investigate what would be the advantages and disadvantages between each method.

‍

The process started by defining the requirements needed to enable a digital engineering tool to produce the expected outputs. Based on our traditional experience, at the start of any project we need to gather a set of input data including location of points to connect, network rules, type of materials, type of surfaces, etc. and cross it with a basis for the design, normally cadastral maps, aerial imagery and/or street view. The requirements are similar in a digital process.

‍

So, what was innovative about this then?

‍

In the traditional way we design the networks in a 2D environment, by the support of aerial imagery. This time however, we were challenged to digitally design the network in a 3D mapping plugin, directly in a set of spherical photos collected by a mobile mapping provider.

‍

‍

Comparing the results by using both approaches we found 3D to be advantageous in the quality of the imagery by being up to date, as well as having the ability to see below roofs, tree crowns, and gaining a higher resolution. Simultaneously, this method has the limitation to see behind walls and obstacles, which we consider to impart constrains in deciding which of type of drilling to use.

‍

Another advantage was the ability to take decisions faster in the perspective of the designer. The fact that the features are designed at the same time the designer "walks” through the street view speeds up the design process. However, the accuracy of the geographical position of the designed features quickly decreases with the distance from the camera, due to its spherical characteristics.

‍

At this point, you may ask why we call it a digital rise, if we still have the designer behind the screen making the decisions. We also asked that question, and we pushed it to the next level by accepting the challenge of evaluating trench lines designed by an AI-based tool, which could identify trees and classify surface types, and consequently make decisions on the geographic placement of the trenches.

‍

The analysis of the results from the tree recognition algorithm, prove they are not yet as accurate as desired, as some trees were not always recognized, and when recognized they were misplaced a couple of meters from the original position.

‍

On the other hand, the classification of distinct types of surfaces was found to be extremely useful and with a great potential to quickly identify and correspond these surface types to the designed features.

The same for the trenches allocating algorithm, as it can help the most to reduce the work in the engineering phase. The places of the street crossing were quite reasonable, and as the algorithm insistently avoided trees and bushes, with the only alternative of trenching in the asphalt, created clearly visual angular deviations helping the designer to easily find and replace these by proper alternatives (tree drilling or other type of drillings).

‍

There are still many possible actions towards another digital rise of the engineering process. At Yungo we are rolling up our sleeves and looking forward to showing you more about it!