In recent years, the Telerobot Consortium of Genova has faced the problem of enabling an autonomous vehicle (SAM) to move in a well known indoor environment. The determination of position and orientation of the robot was solved by integrating odometry, computation of ideal points, and finally the use of annulus-shaped landmarks. These were the by-product of a research that our group accomplished in collaboration with Elsag Bailey (Genova) on the use of projective geometry for the pose estimation of circles in space.
Some years ago, our group envisaged the use of altered Riemannian metrics for path-planning, after an idea of U. Leuthäusser. This Author had used the Riemannian paradigm in an isotropic environment. We exploited it in the richer context of anisotropic situations. Key example was that of a ploughed field. A theoretical extension of the method to strong anisotropy is currently under study.
This opened the way to a collaboration with ENEL-CRA (National Body for Electric Energy - Centre for Research in Automation). Within this project, we are exploring the possibility of extending the method to path- and task-planning of an autonomous vehicle in an agricultural environment.
This is currently done in strict collaboration with Andrea Anghinolfi (who started the whole research as a student in our group) and Luca Costa, both at ALYA (Imola and Modena).