A generic framework for scalable and convergent multi-robot active simultaneous localization, mapping and target tracking

In this paper, a new approach is proposed and analyzed for developing efficient and scalable methodologies for multi-robot active Cooperative Simultaneous Localization And Mapping and Target Tracking (C-SLAMTT). The proposed approach employs an active estimation scheme that switches among linear elements and, as a result, its computational requirements scale linearly with the number of estimated quantities (number of number of robots, landmarks and targets). The parameters of the proposed scheme are calculated off-line using a convex optimization algorithm which is based on Semi Definite Programming (SDP) and approximation using Sum-of-Squares (SoS) polynomials. As shown by rigorous arguments, the estimation accuracy of the proposed scheme is equal to the optimal estimation accuracy plus a term that is inversely proportional to the number of estimator\´s switching elements (or, equivalently, to the memory storage capacity of the robots\´ equipment). The proposed approach can handle various types of constraints such as "stay-within-an-area", obstacle avoidance and maximum speed constraints. The efficiency of the approach is demonstrated on a 3D active cooperative simultaneous mapping and target tracking application employing flying robots.