An informative path planning framework for UAV-based terrain monitoring

Informative Path Planning for Active Field Mapping under Localization Uncertainty

“…Significant recent work has been done on autonomous data gathering strategies in the context of robotics and related fields. The discussion in this section focuses on two main research streams: (1) methods for probabilistic environmental mapping [2,6,9,14], and (2) algorithms for informative planning [1,4,5,12,15,16].…”

“…GPs are a popular non-parameteric Bayesian technique for modeling spatio-temporal phenomena [14]. They have been applied in various active sensing scenarios [1,3,5,16] to gather data based on correlations and uncertainty in continuous maps. However, most of these works assume that the training data for prediction is inherently noise-free, which may lead to inaccuracies if measurements are incorporated at wrong locations and mislead predictive planning algorithms.…”

“…Rapid technological advancements are inciting the use of autonomous mobile robots for exploration and data acquisition. In many marine [1,2], terrestrial [3,4], and airborne [5,6] applications, these systems have the ability to bridge the spatiotemporal divides limiting traditional measurement methods in a safer and more cost-effective manner [7]. However, to fully exploit their potential, algorithms are required for planning efficient informative paths in complex environments under platform-specific constraints.…”

“…This paper examines the problem of active mapping using a robot, where the aim is to recover a continuous 2-D or 3-D field, e.g., of temperature, humidity, etc., using measurements collected by an on-board sensor. In similar setups, most existing strategies [1,5,8] incorrectly assume perfect pose information, which is an implicit requirement for building high-quality maps in initially unknown environments. Our motivation is to improve upon the robustness and accuracy of field reconstructions by allowing the robot to adaptively trade-off between gathering new information (exploration) and maintaining good localization (exploitation).…”

“…3) Evaluations of the approach in a 3-D graph Simultaneous Localization and Mapping (SLAM) setup and a proof of concept in a temperature mapping scenario. We note that our framework can be used in any scalar field mapping scenario, e.g., spatial occupancy [4,6,12], signal strength [2,8], aerial surveillance [5,7], etc., and with any SLAM or localization-only algorithm supplying the pose uncertainty. Specifically, in this work, we focus on a general case where the spatial field information being collected is decoupled from the localization routine.…”