These two robots have opened a new era in I&M of infrastructures, by demonstrating the possibility of installing autonomous machines on (new or old) infrastructures. The Genova San Giorgio Bridge safety comes from four robots, all together protagonists of an extremely innovative application, in the field of civil infrastructures, an absolute novelty at an international level. For the first time, two robots automatically scan the lower surface of the deck, which is generally the most difficult part to reach. In addition, from year to year, the robots will compare the results obtained to predict the onset of any problems and defects. A second pair of Robot-Washes are dedicated to cleaning operations. The task of the Robot-Washes is to clean the wind barriers and solar panels. The Robot-Wash represents a green application since for their maintenance and cleaning activities they will use rain and/or dew as a water resource. Cognitive Mechatronics: the Robots, at any moment, know what is happening around them. Based on the atmospheric conditions, they are able to decide whether they will be able to work or not, whether to interrupt their inspection activity or continue, depending on the assessment of rain intensity or wind. Furthermore, based on the consumption reached by the batteries, they will calculate their actual ability to complete their task or will have to stop at the nearest charging station to recharge. The project was born looking beyond the Genoa bridge. In fact, we are already working on a system that will be functional for all infrastructures, from tunnels to dams, up to skyscrapers, since the roller coasters can also go vertically, as well as horizontally. In practice, there is no limit to the applicability of the new robotic system, with the necessary adaptations.

In addition to the fixed sensors (accelerometers, strain gauges, etc.), visual surveillance is also necessary for the Genoa bridge, where the technicians will be assisted by two pairs of robots who will move on the external and lower parts for the inspection of the viaduct in the less accessible areas (Robot Inspection) and for cleaning of wind barriers and solar panels (Robot Wash). The project of the bridge and the idea of robot for cleaning the sound barriers and solar panels was born from a proposal from Renzo Piano's studio, which contacted the Industrial Robotic Unit (URI) of Italian Institute of Technology. Afterwards, URI created and patented not only the required cleaning robots, but also the first robotic automatic inspection system in the world. This huge project was sponsored by PerGenova (via Fincantieri) and the carried on by a consortium lead by Camozzi Group from Brescia (Italy), specialized in industrial automation, which developed the robots by building the components in carbon fiber (by Ingersoll) and the robot control (by INNSE Berardi). The structural design was carried on by SDA engineering from Padova (Italy), Ubisive from Civitanova Marche (Italy) dealt with the Human Machine Interface and University Politechnique of Marche from Ancona (Italy) developed the algorithm for pattern analysis.

Together with the two machines there are the more than 250 fixed sensors installed by Cetena, sensitive to vibrations and longitudinal, transversal or inclination movements, but Robot Inspection can do much more, having the possibility to move back and forth along the bridge.
It will work like this: when given the green light by a human operator, the robot begins to move along the bridge in a completely autonomous way, on one side and the other, taking about 25 thousand photographs. It takes more or less a week, not because it is slow but because it is attentive: to do its job properly it needs good conditions of light and visibility and if there are none (because it is night, because it rains very hard, because it is cloudy ) stops and waits.
Once the monitoring is complete, the robot compares the result of its inspection with the initial scan, the reference model: it checks point by point, every point where it took a photo, and then evaluates if everything is fine. Evaluate by itself, thanks to the so-called pattern analysis: artificial intelligence allows him to understand if that shadow that is seen at that point can hide a problem or is only due to a cloud, if the oscillation he perceived is a symptom of something more serious or linked to the wind, which in Val Polcevera can blow even over 50 per hour, and so on. In short, it is an AI that does what AIs are very good at: analyzing a huge amount of data and leafing through tens of thousands of photographs, looking for recurrences or anomalies and sending to human colleagues only those that she thinks would need further study.
They are then the ones who will decide whether further information is necessary, whether to mount the aforementioned cameras on Robot Inspection (infrared and 3d) and have it make another pass at the points concerned.