Adaptive Sampling and Online Learning in Multi-Robot Sensor Coverage with Mixture of Gaussian Processes

Abstract Multi-fidelity Gaussian process (GP) modeling is a common approach to employ in resource-expensive computationally demanding algorithms such as optimization, calibration and uncertainty quantification where multiple datasets of varying fidelities are encountered. Briefly, in its simplest form, a multi-fidelity GP is trained on two separate sources of datasets each with its own fidelity level, e.g., a software code/simulator for the low-fidelity source and real-world experiments for the high-fidelity source. Adaptive sampling for multi-fidelity Gaussian processes is a challenging task since we not only seek to estimate the next sampling location of the design variable, but also account for the data fidelities. This issue is often addressed by including the cost of the data sources as an another element in the search criterion in conjunction with an uncertainty reduction metric. In this work, we extent the traditional design of experiment framework for multi-fidelity GPs by partitioning the prediction uncertainty based on the fidelity level and the associated cost of execution. In addition, we utilize the concept of a meta-model believer which quantifies the effect of adding an exploratory design point on the GP uncertainty prediction. We demonstrate the framework using academic examples as well as an industrial application of a steady-state thermodynamic operation point of a fluidized bed process.

Download Full-text

Collective Online Learning of Gaussian Processes in Massive Multi-Agent Systems

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v33i01.33017850 ◽

2019 ◽

Vol 33 ◽

pp. 7850-7857 ◽

Cited By ~ 2

Author(s):

Trong Nghia Hoang ◽

Quang Minh Hoang ◽

Kian Hsiang Low ◽

Jonathan How

Keyword(s):

Online Learning ◽

Gaussian Processes ◽

Message Passing ◽

Transmission Loss ◽

Streaming Data ◽

Multi Agent Systems ◽

Model Fusion ◽

Agent Model ◽

Massive Number ◽

Multi Agent

This paper presents a novel Collective Online Learning of Gaussian Processes (COOL-GP) framework for enabling a massive number of GP inference agents to simultaneously perform (a) efficient online updates of their GP models using their local streaming data with varying correlation structures and (b) decentralized fusion of their resulting online GP models with different learned hyperparameter settings and inducing inputs. To realize this, we exploit the notion of a common encoding structure to encapsulate the local streaming data gathered by any GP inference agent into summary statistics based on our proposed representation, which is amenable to both an efficient online update via an importance sampling trick as well as multi-agent model fusion via decentralized message passing that can exploit sparse connectivity among agents for improving efficiency and enhance the robustness of our framework against transmission loss. We provide a rigorous theoretical analysis of the approximation loss arising from our proposed representation to achieve efficient online updates and model fusion. Empirical evaluations show that COOL-GP is highly effective in model fusion, resilient to information disparity between agents, robust to transmission loss, and can scale to thousands of agents.

Download Full-text

A Multi-Robot Testbed for Adaptive Sampling Experimentation via Radio Frequency Fields

Volume 9: 2015 ASME/IEEE International Conference on Mechatronic and Embedded Systems and Applications ◽

10.1115/detc2015-47697 ◽

2015 ◽

Cited By ~ 1

Author(s):

Jose Acain ◽

Christopher Kitts ◽

Thomas Adamek ◽

Kamak Ebadi ◽

Mike Rasay

Keyword(s):

Scalar Field ◽

Radio Frequency ◽

Adaptive Sampling ◽

Large Scale ◽

Rapid Development ◽

Environmental Data ◽

Sampled Data ◽

Wireless Transmitters ◽

Ongoing Research ◽

Multi Robot

Adaptive navigation is the process by which a vehicle determines where to go based on information received while moving through the field of interest. Adaptive sampling is a specific form of this in which that information is environmental data sampled by the robot. This may be beneficial in order to save time/energy compared to a conventional navigation strategy in which the entire field is traversed. Our work in this area focuses on multi-robot gradient-based techniques for the adaptive sampling of a scalar field. To date, we have experimentally demonstrated multi-robot gradient ascent/descent as well as contour following using automated marine surface vessels. In simulation we have verified controllers for ridge descent / valley ascent as well as saddle point detection and loitering. To support rapid development of our controllers, we have developed a new testbed using wireless transmitters to establish a simple, large-scale, customizable scalar field based on the strength of the radio frequency field. A cluster of six land rovers equipped with radio signal strength sensors is then used to process sampled data, to make adaptive decisions on how to move, and to execute those moves. In this paper, we describe the technical design of the testbed, present initial experimental results, and describe our ongoing research and development work in the area of adaptive sampling and multi-robot control.

Download Full-text

A Trust-based Mixture of Gaussian Processes Model for Reliable Regression in Participatory Sensing

Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2017/540 ◽

2017 ◽

Cited By ~ 2

Author(s):

Qikun Xiang ◽

Jie Zhang ◽

Ido Nevat ◽

Pengfei Zhang

Keyword(s):

Real World ◽

Gaussian Processes ◽

Spatial Regression ◽

Participatory Sensing ◽

Sensing Applications ◽

Different Types ◽

Inaccurate Estimation ◽

Real World Datasets ◽

Gp Model ◽

Mixture Of Gaussian

Data trustworthiness is a crucial issue in real-world participatory sensing applications. Without considering this issue, different types of worker misbehavior, especially the challenging collusion attacks, can result in biased and inaccurate estimation and decision making. We propose a novel trust-based mixture of Gaussian processes (GP) model for spatial regression to jointly detect such misbehavior and accurately estimate the spatial field. We develop a Markov chain Monte Carlo (MCMC)-based algorithm to efficiently perform Bayesian inference of the model. Experiments using two real-world datasets show the superior robustness of our model compared with existing approaches.

Download Full-text