Progetto ROVINA “Robots per l’esplorazione, la conservazione digitale e la visualizzazione dei siti archeologici”

CONCEPT AND OBJECTIVES

Mapping and digitizing archeological sites is an important task to preserve cultural heritage. Systems used for digitizing archeological sites typically build upon static 3D laser scanning technology that has to be brought into archeological sites by humans. While this is acceptable for accessible sites, it prevents the digitization of sites that are inaccessible by humans. In the field of robotics, however, there has recently been a tremendous progress in the development of autonomous robots that offer various services to their users. The task of autonomously exploring three-dimensional underground environments such as caves, however, still opens a series of challenges. The reasons for this are manifold. Such environments pose enormous difficulties to autonomous robots because of the problematic perception (no/poor lighting, unpredictable situations, difficult scene interpretation, complex traversability analysis, . . . ) paired with limited communication capabilities, which restrict the continuous supervision and tele-operation by human users.

The goals of ROVINA are:

• to provide novel technology that supports the preservation of cultural heritage by allowing the acquisition of digital models in hard-to-access environments
• to extend the technology of autonomous navigation for robots designed to explore unknown underground environments such as caves and catacombs
• to develop novel techniques to construct large 3D textured models of these poorly structured environments
• to offer a cost-effective support for performing continuous monitoring of these sites and to enable comparative analysis that will allow to devise better preservation plans.

The technology developed in ROVINA will allow a mobile robot to effectively explore such environments, to acquire a detailed model enriched with semantic information, adapt its navigation behavior according to an enhanced traversability analysis and interact with remote users in an effective way.

To achieve this objective, the ROVINA consortium consists of leading researchers in the fields of digital preservation, autonomous and networked robots, navigation, vision and perception. We will provide novel technologies for the following tasks:

• Flexible user-interfaces providing different levels of autonomy and addressing the needs of different user groups
• Reliable navigation in challenging environments
• Building accurate maps of 3D environments given multi-modal sensory data in an online fashion
• Extracting semantic information from the multi-modal sensor data, i.e., classification of traversability or the identification of archeologically interesting artifacts
• Reliable and cost-effective exploration of unknown and unstructured 3D environments

APPLICATION SCENARIO: THE CATACOMBS OF ROME AND NAPLES

The Catacombs of Rome and Naples have been selected as an application scenario for the project.

At present, only five catacombs are open to the public. The other catacombs are still not open because of their instability. The digitization of these in large parts unexplored and not accessible environments is a challenging endeavor. They are well-suited to serve as as testbed for the ROVINA system for the exploration, digital preservation and explorations of archeological sites.

Operating in such environments requires us to solve several relevant problems simultaneously. This includes the inherent need for robust algorithms due to the inaccessibility of most catacombs, methods that can be executed online since the robot has to act autonomously, sophisticated navigation techniques due to the complexity of the environment which includes debris and dangerous passages, mapping, reconstruction and detection algorithms that can deal with a high degree of repetitive structures in such large-scale environments, user interfaces with a high degree of situation awareness, post processing of the acquired data to allow the detection of differences that occur over time and that allow to determine which regions of the environment need intervention. All this is performed in unknown environments and under potentially unreliable communication links, while requiring a high level of detail in the resulting model. Techniques developed to be successful in these challenging settings can be directly transferred to a large variety of cultural heritage sites.

PARTNERS