MUNINN - Mapping Urban Signage Inventory with Neural Networks in Lviv

MUNINN (Mapping Urban sigNage Inventory with Neural Networks) addresses Lviv’s urgent need for accurate, up-to-date data on its traffic signs and road markings. The city currently relies on outdated paper-based records that hinder effective road safety management and infrastructure maintenance planning. Without a reliable digital inventory, damaged or missing signs go undetected, creating risks for road users and complicating compliance with Ukrainian and European safety standards.
 
LOKI S.r.l., an Italian deep-tech company specialising in AI-driven urban infrastructure monitoring, will deploy its plug-and-play device on a city vehicle to automatically detect, classify and geolocate traffic signs and road markings across Lviv’s central district. The system uses custom-built hardware with camera lenses optimised for sign detection at various heights and angles, combined with AI algorithms trained specifically on Ukrainian signage standards.
 
The pilot follows a two-phase approach. In the first phase (September 2026), a baseline scan is undertaken to map the current state of Lviv’s signage infrastructure. Between phases, the device is handed over to city staff with training. In the second phase (November 2026), city staff operate the system independently, validating the handover model and demonstrating sustainable adoption.
 
All collected data are stored on local city servers, ensuring full municipal ownership and control. The custom visualisation platform, built with Ukrainian language support, provides city planners with an interactive geographic information system (GIS) dashboard featuring near real-time alerts for damaged or missing signs. The project also explores digitising existing paper-based city documentation using AI-powered text and image extraction.