Kriging‐based robotic exploration for soil moisture mapping using a cosmic‐ray sensor

Abstract

Soil moisture monitoring is a fundamental process to enhance agricultural outcomes and to protect the environment. The traditional methods for measuring moisture content in the soil are laborious and expensive, and therefore there is a growing interest in developing sensors and technologies which can reduce the effort and costs. In this work, we propose to use an autonomous mobile robot equipped with a state‐of‐the‐art noncontact soil moisture sensor building moisture maps on the fly and automatically selecting the most optimal sampling locations. We introduce an autonomous exploration strategy driven by the quality of the soil moisture model indicating areas of the field where the information is less precise. The sensor model follows the Poisson distribution and we demonstrate how to integrate such measurements into the kriging framework. We also investigate a range of different exploration strategies and assess their usefulness through a set of evaluation experiments based on real soil moisture data collected from two different fields. We demonstrate the benefits of using the adaptive measurement interval and adaptive sampling strategies for building better quality soil moisture models. The presented method is general and can be applied to other scenarios where the measured phenomena directly affect the acquisition time and need to be spatially mapped.

Error Estimation for Soil Moisture Measurements With Cosmic Ray Neutron Sensing and Implications for Rover Surveys

“…Typically, a larger detector volume improves the counting statistics, and thus reduces the uncertainty of the soil moisture product. The record period of most mobile neutron detectors is between 10 s and 1 min, while typical driving speeds range from 2 to 10 km/h on agricultural fields (Schrön et al, 2018b;Fentanes et al, 2019) to ∼50 km/h for large-scale surveys (e.g., Chrisman and Zreda, 2013;Dong et al, 2014;McJannet et al, 2017;Dong and Ochsner, 2018). In most studies, additional spatial smoothing was applied to the CRN rover measurements by using a temporal moving window filter in order to reduce the uncertainty in the soil moisture estimates (e.g., Schrön et al, 2018b: window size of 3 measurements; Chrisman and Zreda, 2013: window size of 7 measurements).…”

Go with the Flow: Exploration and Mapping of Pedestrian Flow Patterns from Partial Observations

“…Another approach is adopted by next-best-view strategies, which focus on building the initial map based on optimising different criteria. For example, [4] optimises the estimated time to reach a location and the amount of information expected to be gathered there, [5] calculates the entropy decrease in the robot configuration space and uses the estimates for robot mapping, [6], [7] use the Poisson uncertainty of neutron counts to drive the observation and mapping of soil moisture, and [8] presented an information-gain-based exploration approach that takes into account the uncertainty from both the map and the robot's localisation. However, all the above approaches do not attempt to maintain the environment models after their acquisition, meaning that the model will lose accuracy as new changes appear in the environment.…”

Robotic Exploration for Learning Human Motion Patterns

“…Vallvé and Andrade-Cetto [14] propose a method of calculating the entropy decrease in the robot configuration space and then use these estimates to evaluate different exploratory trajectories for robot mapping. Fentanes et al [15] utilise the Poisson uncertainty to drive the observation and mapping of soil moisture by counting neutrons using a cosmic-ray sensor. Stachniss and Burgard [16] presented an information-gain based exploration framework that integrates not only uncertainties of the map, but also the uncertainties of the robot's localisation.…”