At Fiber Connect, IQGeo and Deepomatic demoed how AI is helping fiber providers raise the bar on network quality through automated field validation. In the demo, Thomas Thuillier, GM of North America at Deepomatic, and Jimmy Gagnon, Customer Success Manager at IQGeo, walked through a real-world workflow showing how field technicians can use AI to instantly validate the quality of their work.
Using computer vision models, the platform automatically scans field photos for key visual elements: Are all splices protected? Is the tray dressed properly? Are angles and notches secured? If an issue is detected, it’s flagged directly in the image so the technician can fix it before leaving the site. Once corrected, a new photo is uploaded and confirmed—all while geo-tagging and timestamping every step.
Beyond quality control, Gagnon emphasized the broader value: this process ensures accurate documentation of assets, location, and installation quality. That data feeds directly into a company’s digital twin and fuels smarter network automation. Together, IQGeo’s and Deepomatic’s tools offer a real-time, high-accuracy view of network health—laying the groundwork for faster, more reliable fiber builds.
Thomas Thuillier:
Hi everyone, my name is Thomas and I'm the general manager for North America at Deepomatic.
Jimmy Gagnon:
And I'm Jimmy Gagnon, customer success manager for IQGEO.
Thomas Thuillier:
And what we're going to show you today is how you can use the computer vision technology from Deepomatic to automatically validate the quality of work being performed on the field and enable self-validation. And so what you're seeing here on the screen is first of all the Workflow manager tool from IQGEO, which is a ticketing system. And you can see all the job I have been attributed to as a technician on the field. And when I open the detail of the job, you can see that I have a specific button available to start the validation process through Deepomatic. That setup can be replicated in a whole bunch of different tools. So if you have your own homegrown tool or use a different system, that setup can be replicated. And so now we can see that I have the ability to capture photos from the field of my work.
That photo is going to be automatically analyzed by artificial intelligence model that are going to be looking for specific visual elements in the photo, in order to assess, first of all that I documented my job properly, I took the right photo, it's framed properly, it has all the information that I need. And then inside the photo I will be looking at different visual elements, such as the number of splices, are every splices properly protected with a sleeve? Is the tray properly dressed? Do I have any sharp angle? Is it properly secured under the notches? And you can see here that any little problem that should be corrected will be very clearly flagged by the AI inside the photo, so that I can correct it. Before I leave the site, I retake a photo and I show that I did my job perfectly. On top of that, we capture a lot of information such as the timestamp, the geolocalization, which is really key to document your network efficiently. And Jimmy will tell us a little bit more about why this is important.
Jimmy Gagnon:
Absolutely. So with IQGEO's solutions to cover your entire network lifecycle, the usage of AI is really important, not only to just validate the work that's being done out there, but also to start documenting properly with high accuracy all of your assets. So not only will you know what assets are out there, you'll know exactly where they are and you'll know if they have been installed correctly. So this is actually the foundation to towards a path for digital twin and network automation. So we absolutely had to have a high level of accuracy on the data set, which is provided by the diplomatic solution. So in conjunction with IQGEO, this is really powerful for the organizations to have a real-time visibility on all the assets and where they stand and the quality of the networks. So on this, thank you very much.
Thomas Thuillier:
Thank you, Jimmy. And if you need any more information, need a in-depth demo, don't hesitate to reach out to me, to Jimmy, or anyone else from the IQGEO team and the Deepomatic team. Thank you very much.
Jimmy Gagnon:
Thank you.