Effective Landmark Placement for Robot Indoor Localization With Position Uncertainty Constraints

Abstract

A well-known, crucial problem for indoor positioning of mobile agents (e.g., robots) equipped with exteroceptive sensors is related to the need to deploy reference landmarks in a given environment. Normally, anytime a landmark is detected, an agent estimates its own location and attitude with respect to landmark position and/or orientation in the chosen reference frame. When instead no landmark is recognized, other sensors (e.g., odometers in the case of wheeled robots) can be used to track the agent position and orientation from the last detected landmark. At the moment, landmark placement is usually based just on commonsense criteria, which are not formalized properly.As a result, positioning uncertainty tends to grow unpredictably. On the contrary, the purpose of this paper is to minimize the number of landmarks, while ensuring that localization uncertainty is kept within wanted boundaries. The developed approach relies on the following key features: a dynamic model describing agents' motion, a model predicting the agents' paths within a given environment and, finally, a Conjunctive Normal Form (CNF) formalization of the optimization problem, which can be efficiently (although approximately) solved by a greedy algorithm. The effectiveness of the proposed landmark placement technique is first demonstrated through simulations in a variety of conditions and then it is validated through experiments on the field, by using non-Bayesian and Bayesian position tracking algorithms.

Robot Localisation Using UHF-RFID Tags: A Kalman Smoother Approach †

“…The method has been conceived for industrial scenarios and account for specific position constraints. As an example, the issue related to pallet handling in warehouse applications [ 38 ] requires a pallet-placing uncertainty in the order of a few centimetres with a high level of confidence [ 39 ]. Hence, the main objective of the proposed approach is to robustly reduce to the maximum extent the localisation uncertainty and the vehicle trajectory estimation uncertainty to meet the industrial requirements.…”

A Stacked LSTM-Based Approach for Reducing Semantic Pose Estimation Error

“…The effect of unpredictable nonuniform noise as well as external environmental conditions is also inevitable [35]. To enhance the accuracy of localization, the solutions found in the literature can be classified into: (1) controlling the environment under investigation [36], (2) sensor data fusion [37], [38], (3) improving measurement covariane estimation [39], [30], [35], [40], or (4) correcting measurement errors, which can be further classified into classical [41], [42], [43], [44] and learning approaches [16], [45], [34], [46], [17], [47].…”

“…The work presented in [36] studies the placement of passive tags, used as landmarks, in the environment, to always keep the localization accuracy within a particular range. In another vein, the robustness of indoor localization was supported by accumulating sensory data, which compensated for the limitations of the employed sonar sensors as presented in [37].…”

On-Line Optimal Ranging Sensor Deployment for Robotic Exploration

“…Providing positioning measurements with limited uncertainty for autonomous robot navigation is hard if an absolute reference as the GPS is not available. The achievable positioning performance depends both on the specific technology used by the sensing devices and on the algorithm defined for the placement of such devices [20]. When RF ranging sensors are considered, two different approaches are usually implemented to achieve the optimal placement in an unknown environment: 1) Off-line.…”