RENEW: sensors, artificial intelligence and data for safer roads and more efficient maintenance

Urban roads deteriorate every day, but those responsible for managing them often notice too late. Potholes and cracks are detected through manual visual inspections that are periodic and subjective; more accurate technologies, such as vehicles equipped with lasers and LiDAR, have costs that preclude continuous monitoring. The result is maintenance that chases problems instead of anticipating them.

RENEW (Sensor-based and data-driven framework for safe and efficient PMS) was created to change this approach. The project develops a system that turns urban vehicles into a distributed detection network, analyses the data with artificial intelligence and provides managers with reliable information to decide where, when and how to intervene.

The problem: a PMS that cannot remain a static archive

A Pavement Management System (PMS) is the tool used by municipalities and managing bodies to plan pavement maintenance. Today these systems suffer from several recurring limitations:

The technologies to overcome these limitations already exist, but they are separate: IoT sensors that are cheap but not very sensitive, Computer Vision models not connected to planning systems, analysis tools that do not explain their conclusions. RENEW works precisely on their integration.

Project objectives

RENEW has four main objectives:

Technologies

The device installed on the vehicle: a die-cast aluminium box with IP67 protection, anti-vibration brackets and tamper-proof fixings. It integrates an RGB camera, an inertial measurement unit (accelerometer and gyroscope), a GNSS receiver and an NVIDIA Jetson Orin NX computing module for on-board processing. The firmware is modular and provides for nominal, degraded and offline operating modes.

IAT develops a Tyre Sensing Unit for e-bikes: an accelerometer placed inside the tyre and a pressure and temperature module at the valve, which directly measure the wheel-pavement interaction and derive severity indices.

Processing is split across four levels: sensors acquire data, the Edge node on the vehicle identifies possible defects in real time, the Cloud classifies them and estimates their severity, and the PMS/DSS turns them into planning information. Only significant, georeferenced events leave the vehicle — not continuous video streams.

For on-board detection, lightweight detectors including RF-DETR have been evaluated; for Cloud-side classification, convolutional networks and Vision Transformers. Damage severity (mild, moderate, severe) is estimated by combining images and vibration data.

Every assessment is accompanied by a confidence level and explanations, with an audit log of decisions.

The platform is based on containerised services, device identity management and signed updates, in compliance with GDPR. A georeferenced data warehouse stores the history of events and their provenance.

The partnership

Expected results

By the end of the project, scheduled for January 2027, RENEW will deliver:

The technological objective is to bring the solution from Technology Readiness Level TRL 5 to TRL 7, i.e. to prototype demonstration in an operational environment.

Project status

Launched in August 2025, RENEW has completed the architecture definition, specifications of the Data Collection Device and sensor tyres, design of the software platform and study of AI models. Testing of prototypes, platform development and model training are currently underway.

The first scientific result is the paper "Explainable and Real-Time Road Pothole Analysis using Object Detection Models", presented at Ital-IA 2026, the 6th National Conference on Artificial Intelligence organised by CINI (Rome, 18-19 June 2026).

Impact

For public administrations, RENEW means moving from reactive to data-driven maintenance: intervening before deterioration becomes costly, spending the budget where it is needed, making roads safer for motorists, cyclists and pedestrians. For the Puglia region, it means developing expertise in AI, IoT and Edge Computing applied to infrastructure, born from collaboration between businesses and research.

Project "RENEW – Sensor-based and data-driven framework for safe and efficient PMS" – CUP B67H24001820007 – Practice code XF4UNN8. PR Puglia ERDF ESF+ 2021-2027 (ERDF fund) – Action 1.5 "Interventions for the strengthening of the regional innovation system and support for collaboration between companies and research structures" – "Reti" Call. Project co-funded by the European Union.