Flai presented at 18. PIES 2026, held on 9 and 10 April at the Grand Hotel Portorož. Aleksandar Šašić Kežul of Flycom Technologies and Blaž Vidmar of Flai gave a joint talk, Energetski digitalni dvojček in umetna inteligenca, on the role of LiDAR data and artificial intelligence in building energy digital twins.
PIES (Posvetovanje informatikov v energetiki Slovenije) is one of the main conferences for the IT community in the Slovenian energy sector. This edition was built around how AI supports digitalization in the operation, planning and maintenance of energy systems, which is where the LiDAR side of the problem starts to matter.
Where classification fits in an energy digital twin
A digital twin of a grid is only useful if it holds the grid as objects rather than as an undifferentiated cloud of points. That is the step the talk focused on: taking a corridor survey and separating it into the things an asset manager actually reasons about.
Conductors, lattice towers and wooden poles come out as separate classes, structure by structure, along the whole corridor. Once the infrastructure is separated that way, the analyses that utilities care about follow from it: vegetation encroachment along the corridor, clearance checks, and per-structure asset records. More on that in power line classification.
Thanks to the organizers for the invitation.