Javier G. Monroy

The Multi-Chamber Electronic Nose--an improved olfaction sensor for mobile robotics.

One of the major disadvantages of the use of Metal Oxide Semiconductor (MOS) technology as a transducer for electronic gas sensing devices (e-noses) is the long recovery period needed after each gas exposure. This severely restricts its usage in applications where the gas concentrations may change rapidly, as in mobile robotic olfaction, where allowing for sensor recovery forces the robot to move at a very low speed, almost incompatible with any practical robot operation. This paper describes the design of a new e-nose which overcomes, to a great extent, such a limitation. The proposed e-nose, called Multi-Chamber Electronic Nose (MCE-nose), comprises several identical sets of MOS sensors accommodated in separate chambers (four in our current prototype), which alternate between sensing and recovery states, providing, as a whole, a device capable of sensing changes in chemical concentrations faster. The utility and performance of the MCE-nose in mobile robotic olfaction is shown through several experiments involving rapid sensing of gas concentration and mobile robot gas mapping.

Overcoming the Slow Recovery of MOX Gas Sensors through a System Modeling Approach

Metal Oxide Semiconductor (MOX) gas transducers are one of the preferable technologies to build electronic noses because of their high sensitivity and low price. In this paper we present an approach to overcome to a certain extent one of their major disadvantages: their slow recovery time (tens of seconds), which limits their suitability to applications where the sensor is exposed to rapid changes of the gas concentration. Our proposal consists of exploiting a double first-order model of the MOX-based sensor from which a steady-state output is anticipated in real time given measurements of the transient state signal. This approach assumes that the nature of the volatile is known and requires a precalibration of the system time constants for each substance, an issue that is also described in the paper. The applicability of the proposed approach is validated with several experiments in real, uncontrolled scenarios with a mobile robot bearing an e-nose.

Time-variant gas distribution mapping with obstacle information

This paper addresses the problem of estimating the spatial distribution of volatile substances using a mobile robot equipped with an electronic nose. Our work contributes an effective solution to two important problems that have been disregarded so far: First, obstacles in the environment (walls, furniture,...) do affect the gas spatial distribution. Second, when combining odor measurements taken at different instants of time, their ‘ages’ must be taken into account to model the ephemeral nature of gas distributions. In order to incorporate these two characteristics into the mapping process we propose modeling the spatial distribution of gases as a Gaussian Markov random field. This mathematical framework allows us to consider both: (i) the vanishing information of gas readings by means of a time-increasing uncertainty in sensor measurements, and (ii) the influence of objects in the environment by means of correlations among the different areas. Experimental validation is provided with both, simulated and real-world datasets, demonstrating the out-performance of our method when compared to previous standard techniques in gas mapping.

Calibration of MOX gas sensors in open sampling systems based on Gaussian Processes

Calibration of metal oxide (MOX) gas sensors for continuous monitoring is a complex problem due to the highly dynamic characteristics of the gas sensor signal when exposed to a natural environment in an Open Sampling System (OSS). This work presents a probabilistic approach to the calibration of MOX gas sensors using Gaussian Processes (GP). The proposed approach estimates for every sensor measurement a probability distribution of the corresponding gas concentration, which enables the calculation of confidence intervals for the predicted concentrations. Being able to predict the uncertainty about the concentration related to a particular sensor response is particularly advantageous in OSS applications where typically many sources of uncertainty exist. The proposed approach has been tested with an experimental setup where an array of MOX sensors and a Photo Ionization Detector (PID) are placed downwind with respect to the gas source. The PID is used to obtain ground truth concentration measurements. Comparison with standard calibration methods demonstrate the advantage of the proposed approach.

Odor recognition in robotics applications by discriminative time-series modeling

Odor classification by a robot equipped with an electronic nose (e-nose) is a challenging task for pattern recognition since volatiles have to be classified quickly and reliably even in the case of short measurement sequences, gathered under operation in the field. Signals obtained in these circumstances are characterized by a high-dimensionality, which limits the use of classical classification techniques based on unsupervised and semi-supervised settings, and where predictive variables can be only identified using wrapper or post-processing techniques. In this paper, we consider generative topographic mapping through time (GTM-TT) as an unsupervised model for time-series inspection, based on hidden Markov models regularized by topographic constraints. We further extend the model such that supervised classification and relevance learning can be integrated, resulting in supervised GTM-TT. Then, we evaluate the suitability of this new technique for the odor classification problem in robotics applications. The performance is compared with classical techniques as nearest neighbor, as an absolute baseline, support vector machine and a recent time-series kernel approach, demonstrating the eligibility of our approach for high-dimensional data. Additionally, we exploit the learning system introduced in this work, providing a measure of the relevance of each sensor and individual time points in the classification process, from which important information can be extracted.

A Kalman filter based approach to probabilistic gas distribution mapping

Building a model of gas concentrations has important industrial and environmental applications, and mobile robots on their own or in cooperation with stationary sensors play an important role in this task. Since an exact analytical description of turbulent flow remains an intractable problem, we propose an approximate approach which not only estimates the concentrations but also their variances for each location. Our point of view is that of sequential Bayesian estimation given a lattice of 2D cells treated as hidden variables. We first discuss how a simple Kalman filter provides a solution to the estimation problem. To overcome the quadratic computational complexity with the mapped area exhibited by a straighforward application of Kalman filtering, we introduce a sparse implementation which runs in constant time. Experimental results for a real robot validate the proposed method.

GADEN : A 3D Gas Dispersion Simulator for Mobile Robot Olfaction in Realistic Environments

This work presents a simulation framework developed under the widely used Robot Operating System (ROS) to enable the validation of robotics systems and gas sensing algorithms under realistic environments. The framework is rooted in the principles of computational fluid dynamics and filament dispersion theory, modeling wind flow and gas dispersion in 3D real-world scenarios (i.e., accounting for walls, furniture, etc.). Moreover, it integrates the simulation of different environmental sensors, such as metal oxide gas sensors, photo ionization detectors, or anemometers. We illustrate the potential and applicability of the proposed tool by presenting a simulation case in a complex and realistic office-like environment where gas leaks of different chemicals occur simultaneously. Furthermore, we accomplish quantitative and qualitative validation by comparing our simulated results against real-world data recorded inside a wind tunnel where methane was released under different wind flow profiles. Based on these results, we conclude that our simulation framework can provide a good approximation to real world measurements when advective airflows are present in the environment.

A configurable smart e-nose for spatio-temporal olfactory analysis

This paper describes a novel electronic nose (e-nose) aimed at applications that require knowing not only the gas composition and concentration, but also its temporal and spatial evolution. This is done by capturing additional information related to the chemical substance such as the time-stamp and geo-location of the measurements, as well as other physical magnitudes of the environment like temperature and humidity for correcting and interpreting the data. The device has been conceived following a modular architecture as a set of independent smart modules, which are interconnected and controlled through an I2C interface by a central processing unit. Each smart module can identify itself, store settings for auto-configuration and perform signal pre-processing of the measured variables. Smart module types include: chemical sensors, communication interfaces, batteries, data storage, GPS, temperature and humidity.

Monitoring household garbage odors in urban areas through distribution maps

This paper presents an experimental study of the suitability of a mobile e-nose (carried on a bike) for the monitoring of unpleasant and potentially harmful odors in urban areas, likely coming from residential waste containers. The objective is to obtain a spatial and temporal representation of such odors by means of a gas distribution map, from which valuable information such as the location, or the time-intervals of maximum strength of the nuisance odors can be inferred. As a case of study, the results of a monitoring campaign carried out in a town in southern Spain are presented. The campaign comprises nine measurement runs distributed along three consecutive days, with a total path of more than 90Km. Upon the results, it is concluded the feasibility and potential of the approach, but also the need for a precise sensor characterization and for gas classification.

A robotic experiment toward understanding human gas-source localization strategies

This paper describes an experiment for gas-source localization with a human-teleoperated mobile robot devised to gather data on how humans search for odor-sources. To that end, more than 150 repetitions of the search process are recorded for 69 test subjects, under 4 sensor configurations (including electronic nose, anemometer and video camera) and 4 scenarios (i.e. with different wind-flow conditions and gas-source position). The experiment has been carried out with a ROS-based simulator that allows driving the robot while recording data of interest (e.g. driving commands, robot localization, sensor measurements, ground-truth, etc.) for further analyzing the human process of gas-source searching, and computational fluid dynamics (CFD) to generate realistic and repeatable test conditions. The manuscript describes the different environmental parameters and sensor combinations of the experiment, and explains the methodology under which it was executed. The obtained dataset is publicly available at http://mapir.isa.uma.es/mapirwebsite/index.php/253-gsl-dataset.