Active Localization of Gas Leaks Using Fluid Simulation

Abstract

Sensors are routinely mounted on robots to acquire various forms of measurements in spatio-temporal fields. Locating features within these fields and reconstruction (mapping) of the dense fields can be challenging in resource-constrained situations, such as when trying to locate the source of a gas leak from a small number of measurements. In such cases, a model of the underlying complex dynamics can be exploited to discover informative paths within the field. We use a fluid simulator as a model, to guide inference for the location of a gas leak. We perform localization via minimization of the discrepancy between observed measurements and gas concentrations predicted by the simulator. Our method is able to account for dynamically varying parameters of wind flow (e.g., direction and strength), and its effects on the observed distribution of gas. We develop algorithms for off-line inference as well as for on-line path discovery via active sensing. We demonstrate the efficiency, accuracy and versatility of our algorithm using experiments with a physical robot conducted in outdoor environments. We deploy an unmanned air vehicle (UAV) mounted with a CO2 sensor to automatically seek out a gas cylinder emitting CO2 via a nozzle. We evaluate the accuracy of our algorithm by measuring the error in the inferred location of the nozzle, based on which we show that our proposed approach is competitive with respect to state of the art baselines.

GadenTools: A Toolkit for Testing and Simulating Robotic Olfaction Tasks with Jupyter Notebook Support

“…However, because CFD simulations with state-of-the-art numerical solvers are difficult to set up and offer no integration with robotics software, many researchers end up opting for custom-made simulation tools that are more appropriate for robotic olfaction [2,13]. Since these tools are not commercial software meant to be used by the non-expert public, even when they are available online they are often difficult for new users to interact with and lack proper documentation.…”

“…As a result, GadenTools allows for a fast prototyping and development of contents related to MRO. All the resources, including the source code, are made available in an online open repository 2 .…”

Vid2Param: Modeling of Dynamics Parameters From Video

“…When simulation calls are expensive, a prior over the parameter space can be enforced, e.g. Gaussian Processes, and Bayesian Optimization can be used [42] [41] [34] [5]. Our approach is closely related to [52] as we use supervision during the training phase of our model, and then use this learned approximation at test time.…”

Bayesian Optimisation for Active Monitoring of Air Pollution

“…Bayesian optimisation (BO) has been used for pollution monitoring previously, but has tended either to estimate hyperparameters or their distributions in ways that do not generalise to new sites in a sample-efficient way (Ainslie et al 2009) or to focus on robot trajectories (Morere, Marchant, and Ramos 2017;Singh et al 2010;Marchant and Ramos 2012), a distinct problem from the one we consider here. Work has also been done on monitoring gas leaks (Reggente and Lilienthal 2009;Asenov et al 2019), but again by using a single moving agent. An exception is Hellan, Lucas, and Goddard (2020), but it is limited to satellite data as a proxy for ground measurements.…”