UAVs and protection of Cultural Heritage sites during emergency situations


During the 2016 Central Italy earthquake, drones have been extensively used by the Italian National Fire Service (CNVVF) to protect Cultural Heritage buildings. In particular, their use allowed firefighters to acquire important data about the conditions of the building without exposing themselves to risk of sudden structural collapse due to earthquakes. The image shows a typical night scenario of the earthquake. (image credits: CNVVF)

Being aware of the situation is one of the most important goals that emergency services need when they design the systems and the procedures to be used during or in the aftermath of a disaster. Situation awareness has many different aspects and needs a flow of information (possibly) in real time from a wide variety of data sources. Such data feed the systems that let emergency managers to assess the situation and take their decisions.

In this framework, the research and the end-user’s needs in the field of Cultural Heritage protection are aiming to integrated systems, featuring sensors and state-of-the-art platforms that have to be built in order to offer the needed information about the conditions of artefacts and the damages they’ve suffered for any kind of natural or man-made reason. According such strategy, heterogeneous and distributed data sources should communicate among the main system, generating a flow of data and information through the traditional internet channel. In this framework, sensors infrastructure based on UAV for surveying, diagnosis and monitoring open-space Cultural Heritage sites could be part of a system that would need technologies and innovative approaches to recognise images (collected by UAVs) along with models and techniques of information fusion.

The steel lattice built to protect the remaining parts of the San Benedetto church in Norcia from the collapse. Data needed to design it have been acquired using drones. (image credits: CNVVF)

Exploiting complex event processing techniques and technologies, the extracted information and/or the deducted/determined domain events, would be aggregated and correlated each other in order to bring out potential dangerous or critical situations, ranging from the recognition, validation and localization of signals and events that may suggest the need for monitoring, surveying or warning for disaster prevention, assessing the level of risk (Surveillance & Monitoring Services, Surveying & Diagnosis Services, Quick Damage Assessment Services).

A case study: the 2016 earthquake in Central Italy

In the 2016 earthquake in central Italy an increasing use of drones operated by Italian firefighters (CNVVF) has been recorded, from the early stages of the emergency, in order to have a quick and detailed overview of the magnitude of the damage suffered by major historical and artistic buildings. Such activity has been carried out in the framework of the new procedures adopted to secure buildings damaged in large scale emergency.

In case of earthquake, buildings can pose severe safety problems when damaged. Drones allow acquire data from the inside or in normally not accessible parts without adding risks to firefighters (image credits: CNVVF)

The same tools were used to define the urban areas with the highest number of building collapses. The drones, equipped with  instrumentation for the photographic survey, have allowed the acquisition of a quantity of gigabytes of high-resolution images of the state of post seismic event locations. In particular, the flight of drones helped to identify the state of damage of all the historic buildings and churches of great artistic importance, located in the red area or not allowed area. These data analysis was significant in order to assess the real risk of further collapses and to design effective shoring systems to support unsafe parts still standing.

The aerial photogrammetric data obtained with several daily sorties of drones, are served by specific input software for rapid return and creation of 3D models, or integrated with cadastral data and geomorphological were a valuable support for the knowledge of the actual operating environment where the teams of firefighters intervened for the search and rescue people. In addition, this post processing has enabled, at the end of the rescue of the population, even a more accurate assessment of the damage and consequently a cost estimate as early as the early stages of the emergency.

Obviously, the accuracy of the data obtained (eg. point clouds, surface models and orthophotos) is not comparable with other system such as LIDAR, however, it represents a valid activity rescue tool support allowing to achieve a good evaluation of the severity of the scenario, and then an estimate of the timing necessary for the refurbishment of the primary infrastructure such as roads, electrical networks etc..

In the specific context, the Italian Fire Corps (CNVVF) special units experts in topography during rescue operations (and able to initiate the procedures for mapping), have scoured the areas affected by the quake. The VHF  radio network of the CNVVF (equipped with GPS module and interfaced to specific software on tablet for tracking and geo-referencing), has let them to prepare maps where the information gathered from multiple sources, were processed by experts in GIS systems and transformed it in shapefiles or other formats widely used on platforms such as Google Maps. In this kind of scenarios, the activities needed to assess and restore safety of historic or cultural buildings can be supported by the research as the one carried out in the H2020 STORM project. The task of assessing quickly and in safety condition the damages suffered by historical or cultural buildings has brought to a wide use of UAVs by the CNVVF in the 2016 earthquake. The images recorded by the sensors that have equipped UAVs have been useful to emergency tasks, but their utility would be boosted by the comparison between data detected by LIDAR before and after the disaster event. The STORM pilots scenarios are aiming at integrating UAVs, LIDAR images and procedures shared between cultural heritage managers and CNVVF, in order to let them assess on the scenario and with the best possible resolution the damages a natural event has caused to buildings.

A paper concerning  the use of drones (STORM project and the use of UAV to improve emergency management of  disasters threatening cultural heritage), presented in the UAV&SAR2017 (Rome, 29th March, 2017) Workshop  can be downloaded here: Guerrieri Marsella STORM_UAVSAR_def (1)

LiDAR and Cultural Heritage in emergency situations: the H2020 STORM project first outcomes

The possibility of acquiring images and data of damaged buildings during the first phases of the emergencies is crucial to put them in safe conditions. (Image credits: CNVVF National Fire and Rescue Services  – Italy)

The project STORM (Safeguarding Cultural Heritage through Technical and Organisational Resources Management) has been funded by the Horizon 2020 EU Program and aims at defining a platform that managers of cultural heritage sites can use in improving preparedness, managing emergencies and planning restoration of damaged buildings.

The project specifically considers risks that the cultural sites have to face from either long-term degradation (whose action is far slower than the typical applications of feedback controls), or extreme traumatic events (whose action is much faster). Their common nature is the climate change. So, the specific scope of the project is creating a technological platform that allows a systematic comparison between a real (measured) state and a desired theoretical state.

Assumptions are kept to the minimum possible level and the difference (the measured error signal), is the main input for whatever algorithm may be used to compute the action (input) that needs to be applied to the mitigation process to achieve the desired objective. So, in other words, reliable and up-to-date measures of the key risk variables are the base line for the STORM predictive model but also for the identification of better intervention actions in terms of restoration and conservation of original materials that will be the starting point for a long term mitigation strategies. As a consequence, needs take into account the use of a large number of sensors, in order to acquire the most useful data. For example, in the case of a progressive relative displacement of a structural beam of an ancient monument, over time comparison of periodical LIDAR based detection of the artefact overall 3D model can be used to detect the small differences in the beam’s position over time.

UAVs have been extensively used by the Italian National Fire Service during the 2016 Central Italy Earthquake to survey damages suffered by Cultural Heritage Buildings. (Image credits: CNVVF National Fire and Rescue Services  – Italy)

What is a LiDAR?

According Wikipedia, Lidar (also called LIDAR, LiDAR, and LADAR) is a surveying method that measures distance to a target by illuminating that target with a laser light. The name lidar, sometimes considered an acronym of Light Detection And Ranging (sometimes Light Imaging, Detection, And Ranging), was originally a portmanteau of light and radar. Lidar is popularly used to make high-resolution maps, with applications in geodesy, geomatics, archaeology, geography, geology, geomorphology, seismology, forestry, atmospheric physics, laser guidance, airborne laser swath mapping (ALSM), and laser altimetry. Lidar sometimes is called laser scanning and 3D scanning, with terrestrial, airborne, and mobile applications.

General schema of LiDAR. (Image credits: Wikipedia)

How Cultural Heritage can benefit of LiDAR (according STORM Project)

Based on such information a team of experts (structural engineers, archaeologists, geologists, restorers) will cooperate, in order to understand the causes and find the most adequate response. In this example, the action cannot be predetermined (nor taken automatically of course), but instead requires a careful and accurate cooperative design and planning of the action in order for it to be as effective and as unobtrusive as possible.
When a disaster occurs, general guidelines related to a wide range of events (e.g. flood, earthquake), existing for the specific site, must be dynamically adapted in near real time by ad-hoc team of experts in order to identify the most urgent recovery actions for the specific emergency. So, LIDAR sensors used for structural evaluation and track-changes of the artefact in terms of erosion monitoring as also for geomorphological assessment and mapping of the protected area can offer a valuable support to managers. Moreover, photogrammetric reconstruction by means of historical and contemporary aerial photography to track-changes can support when it comes to assessing the damages through time and forecast potential future threats

LIDAR equipment have been used until now mostly on movable supports, that are steadily placed on the ground to let an accurate record of data. More recently, RPAS devices have been tested as platform to be equipped with regular camera (high resolution RGB still pictures) for monitoring and mapping, Near Infrared camera and thermal and multispectral sensors or the localization and monitoring of buried structures, light-weight LiDAR for higher resolution 3D scanning. Such possibility has demonstrate its extreme importance during emergency situations: in fact, accessing parts of buildings in some cases can be difficult or can pose a severe risk to rescuers. During the rescue operations of the Central Italy earthquake of August 2016, RPAS mounted LIDAR have been used in many scenarios by the Italian National Fire Service and a complete report of such use hasn’t been published yet.

In which scenarios can LIDAR sensors prove to give data not replaceable by other sensors or any operational procedures? One of the first case is any natural or man-made threat that can damage the structures of heritage buildings. Suppose that, after an earthquake, in an ancient masonry buildings fixtures are identified. Even if, in general, it is possible to track the evolution of a fixture in a building, in the larger buildings it is actually impossible to be certain that a damage has been produced by a specific event.

The Italian pilot site of STORM will be tested in the Diocletian Baths complex, in Rome,  a small portion of which is showed in the image. (Image credits: Wikipedia)

It could have been caused previously for any reason (i.e. failure of foundation). The answer that the Italian STORM pilot site of museum of Terme di Diocleziano (Diocletian Baths – Rome) is currently testing is based on a LIDAR scanner of the buildings.

The hypothetical scenario sees a rescue call to firefighters that arrive with their own LIDAR, scan the portion of the building damaged and compare their results with the data previously acquired by the museum managers. As it’s known LiDAR needs time and, mostly, large quantity of data storage, but a small portion of a building is much more manageable. So, even with a high definition setting, the procedure could offer a new possibility to improve the reliability of the assessment that rescuers have to do during  operations.

Dr. Ing. Stefano Marsella (CNVVF) for STORM Project