@conference{jinicassp2013,

title = {Active analysis of chemical mixtures with multi-modal sparse non-negative least squares},

author = {J Huang and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/jinicassp2013.pdf},

year  = {2013},

date = {2013-02-28},

booktitle = {38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP)},

keywords = {Active sensing, Chemical sensors, Infrared spectroscopy},

pubstate = {published},

tppubtype = {conference}

}

@conference{gutierrez2011invited,

title = {Invited: Advances in Active and Adaptive Chemical Sensing},

author = {R Gutierrez-Osuna and R Gosangi and A Hierlemann},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/gutierrez2011invited.pdf},

year  = {2011},

date = {2011-01-01},

booktitle = {Proceedings of the 14th International Symposium on Olfaction and Electronic Nose},

abstract = {In this presentation, we will review advances at the chemical sensor and data processing levels that may enable the development of adaptive sensors. We will briefly discuss techniques at the sensor and system levels, including modulation of internal parameters (e.g., operating temperatures and absorption wavelengths) and external parameters (e.g., exposure times, preconcentration temperatures). At the signal processing level, we will overview adaptive filtering strategies that may be used to cancel interferences (e.g., environmental variables, drift), adaptive classification techniques for incremental learning in dynamic environments, and active sensing methods for on‐line optimization of sensor arrays and individual tunable sensors. We will also present a methodology based on probabilistic graphical models that may be used to model the dynamic response of metal‐oxide sensors under temperature modulation and select suitable temperature sequences on‐the‐fly, as the sensor interacts with its environment.},

keywords = {Active sensing},

pubstate = {published},

tppubtype = {conference}

}

@conference{gosangi2011quantification,

title = {Quantification of Gas Mixtures with Active Recursive Estimation},

author = {R Gosangi and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/gosangi2011quantification.pdf},

year  = {2011},

date = {2011-01-01},

booktitle = {Proceedings of the 14th International Symposium on Olfaction and Electronic Nose},

journal = {AIP Conference Proceedings},

pages = {23-24},

abstract = {We present an active‐sensing strategy to estimate the concentrations in a gas mixture using temperature modulation of metal‐oxide (MOX) sensors. The approach is based on recursive Bayesian estimation and uses an information‐theoretic criterion to select operating temperatures on‐the‐fly. Recursive estimation has been widely used in mobile robotics, e.g., for localization purposes. Here, we employ a similar approach to estimate the concentrations of the constituents in a gas mixture. In this formulation, we represent a concentration profile as a discrete state and maintain a ‘belief’ distribution that represents the probability of each state. We employ a Bayes filter to update the belief distribution whenever new sensor measurements arrive, and a mutual‐information criterion to select the next operating temperature. This allows us to optimize the temperature program in real time, as the sensor interacts with its environment. We validate our approach on a simulated dataset generated from temperature modulated responses of a MOX sensor exposed to a mixture of three analytes. The results presented here provide a preliminary proof of concept for an agile approach to quantifying gas mixtures.},

keywords = {Active sensing, Metal-oxide sensors},

pubstate = {published},

tppubtype = {conference}

}

@conference{huang2011active,

title = {Active Sensing with Fabry-Perot Infrared Interferometers},

author = {J Huang and R Gosangi and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/huang2011active.pdf},

year  = {2011},

date = {2011-01-01},

booktitle = {Proceedings of the 14th International Symposium on Olfaction and Electronic Nose},

journal = {AIP Conference Proceedings},

pages = {31-32},

abstract = {In this article, we describe an active‐sensing framework for infrared (IR) spectroscopy. The goal is to generate a sequence of wavelengths that best discriminates among chemicals. Unlike feature‐selection strategies, the sequence is selected on‐the‐fly as the device acquires data. The framework models the problem as a Partially Observable Markov Decision Process (POMDP), which is solved by a greedy myopic algorithm. In previous work [1], we had applied this framework to temperature‐modulated metal oxide sensor. Here, we adapt the framework to a tunable IR sensor based on Fabry‐Perot interferometers (FPI). FPIs provide a low‐cost alternative to traditional Fourier Transform Infrared Spectroscopy (FTIR), though at the expense of a narrower effective range and lower spectral resolution. Here, we first test whether the framework can scale up to large problems consisting 27 chemicals with 60 dimensions; our previous work on metal oxide sensors employed three chemicals and 7 dimensions. For this purpose, FPI spectra are simulated from FTIR. Then we validate the framework experimentally on 3 chemicals using a prototype instrument based on FPIs. These preliminary results are encouraging and indicate that the framework is able to solve classification problems of reasonable size.},

keywords = {Active sensing, Infrared spectroscopy},

pubstate = {published},

tppubtype = {conference}

}

@conference{gosangi2011data,

title = {Data-driven Modeling of Metal-oxide Sensors with Dynamic Bayesian Networks},

author = {R Gosangi and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/gosangi2011data.pdf},

year  = {2011},

date = {2011-01-01},

booktitle = {Proceedings of the 14th International Symposium on Olfaction and Electronic Nose},

abstract = {We present a data‐driven probabilistic framework to model the transient response of MOX sensors modulated with a sequence of voltage steps. Analytical models of MOX sensors are usually built based on the physico‐chemical properties of the sensing materials. Although building these models provides an insight into the sensor behavior, they also require a thorough understanding of the underlying operating principles. Here we propose a data‐driven approach to characterize the dynamical relationship between sensor inputs and outputs. Namely, we use dynamic Bayesian networks (DBNs), probabilistic models that represent temporal relations between a set of random variables. We identify a set of control variables that influence the sensor responses, create a graphical representation that captures the causal relations between these variables, and finally train the model with experimental data. We validated the approach on experimental data in terms of predictive accuracy and classification performance. Our results show that DBNs can accurately predict the dynamic response of MOX sensors, as well as capture the discriminatory information present in the sensor transients.},

keywords = {Active sensing, Metal-oxide sensors},

pubstate = {published},

tppubtype = {conference}

}

@conference{gosangi2010sensorsc,

title = {Energy-aware active chemical sensing},

author = {R Gosangi and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/gosangi2010sensorsc.pdf},

year  = {2010},

date = {2010-01-01},

booktitle = {Proceedings of IEEE Sensors},

pages = {1094--1099},

organization = {IEEE},

abstract = {We propose an adaptive sensing framework for metal-oxide (MOX) sensors that seeks to minimize energy consumption through temperature modulation. Our approach generates temperature programs by means of an active-sensing strategy combined with an objective function that penalizes power consumption. The problem is modeled as a partially observable Markov decision process (POMDP) and solved with a myopic policy that operates in real time. The policy selects sensing actions (i.e., temperature pulses) that balance misclassification costs (e.g., chemicals identified as the wrong target) and sensing costs (i.e., power consumption). We experimentally validate the method on a ternary chemical discrimination problem, and compare it against a "passive classifier." Our results show that, for a given energy budget, the active-sensing strategy selects temperatures with more discriminatory information than those of the passive classifier by penalizing pulses of higher temperature and longer durations.},

keywords = {Active sensing, Chemical sensors, Metal-oxide sensors, Temperature modulation},

pubstate = {published},

tppubtype = {conference}

}

@conference{gosangi2009active,

title = {Active chemical sensing with partially observable Markov decision processes},

author = {R Gosangi and R Gutierrez-Osuna},

url = {https://psi.engr.tamu.edu/wp-content/uploads/2018/01/gosangi2009active.pdf},

year  = {2009},

date = {2009-01-01},

booktitle = {Proceedings of 13th International Symposium on Olfaction and Electronic Noses},

abstract = {We present an active‐perception strategy to optimize the temperature program of metal‐oxide sensors in real time, as the sensor reacts with its environment. We model the problem as a partially observable Markov decision process (POMDP), where actions correspond to measurements at particular temperatures, and the agent is to find a temperature sequence that minimizes the Bayes risk. We validate the method on a binary classification problem with a simulated sensor. Our results show that the method provides a balance between classification rate and sensing costs.},

keywords = {Active sensing},

pubstate = {published},

tppubtype = {conference}

}