Filtern
Erscheinungsjahr
Dokumenttyp
- Beitrag zu einem Tagungsband (66)
- Vortrag (43)
- Posterpräsentation (41)
- Zeitschriftenartikel (34)
- Beitrag zu einem Sammelband (3)
- Sonstiges (2)
- Forschungsbericht (2)
- Dissertation (1)
Sprache
- Englisch (140)
- Deutsch (48)
- Mehrsprachig (3)
- Polnisch (1)
Schlagworte
- Mobile Robot Olfaction (26)
- Nano aerial robot (19)
- Gas source localization (14)
- TDLAS (13)
- Tomographic reconstruction of gas plumes (13)
- Gas storage areas (12)
- Membrane-based gas sensing (12)
- Subsurface monitoring (12)
- Swarm (12)
- Tunable Diode Laser Absorption Spectroscopy (TDLAS) (12)
Organisationseinheit der BAM
- 8 Zerstörungsfreie Prüfung (96)
- 8.1 Sensorik, mess- und prüftechnische Verfahren (96)
- 2 Prozess- und Anlagensicherheit (13)
- 2.1 Sicherheit von Energieträgern (12)
- 1 Analytische Chemie; Referenzmaterialien (9)
- 1.5 Proteinanalytik (6)
- 3 Gefahrgutumschließungen; Energiespeicher (4)
- 3.0 Abteilungsleitung und andere (3)
- 8.2 Zerstörungsfreie Prüfmethoden für das Bauwesen (3)
- 1.9 Chemische und optische Sensorik (2)
Paper des Monats
- ja (1)
Setting up sensors for the purpose of environmental monitoring should be a matter of days, but often drags over weeks or even months, preventing scientists from doing real research. Additionally, the newly developed hardware and software solutions are often reinventing existing wheels. In this short paper, we revisit the design of our environmental sensing node that has been monitoring industrial areas over a span of two years. We share our findings and lessons learned. Based on this, we outline how a new generation of sensing node(s) can look like.
This paper outlines significant advancements in our previously developed aerial gas tomography system, now optimized to reconstruct 2D tomographic slices of gas plumes with enhanced precision in outdoor environments. The core of our system is an aerial robot equipped with a custom-built 3-axis aerial gimbal, a Tunable Diode Laser Absorption Spectroscopy (TDLAS) sensor for CH4 measurements, a laser rangefinder, and a wide-angle camera, combined with a state-of-the-art gas tomography algorithm. In real-world experiments, we sent the aerial robot along gate-shaped flight patterns over a semi-controlled environment with a static-like gas plume, providing a welldefined ground truth for system evaluation. The reconstructed cross-sectional 2D images closely matched the known ground truth concentration, confirming the system’s high accuracy and reliability. The demonstrated system’s capabilities open doors for potential applications in environmental monitoring and industrial safety, though further testing is planned to ascertain the system’s operational boundaries fully.
Gas Distribution Mapping (GDM) is essential in monitoring hazardous environments, where uneven sampling and spatial sparsity of data present significant challenges. Traditional methods for GDM often fall short in accuracy and expressiveness. Modern learning-based approaches employing Convolutional Neural Networks (CNNs) require regular-sized input data, limiting their adaptability to irregular and sparse datasets typically encountered in GDM. This study addresses these shortcomings by showcasing Graph Neural Networks (GNNs) for learningbased GDM on irregular and spatially sparse sensor data. Our Radius-Based, Bi-Directionally connected GNN (RABI-GNN) was trained on a synthetic gas distribution dataset on which it outperforms our previous CNN-based model while overcoming its constraints. We demonstrate the flexibility of RABI-GNN by applying it to real-world data obtained in an industrial steel factory, highlighting promising opportunities for more accurate GDM models.
Gas Distribution Mapping (GDM) is a valuable tool for monitoring the distribution of gases in a wide range of applications, including environmental monitoring, emergency response, and industrial safety. While GDM is actively researched in the scope of gas-sensitive mobile robots (Mobile Robot Olfaction), there is a potential for broader applications utilizing sensor networks. This study aims to address the lack of deep learning approaches in GDM and explore their potential for improved mapping of gas distributions. In this paper, we introduce Gas Distribution Decoder (GDD), a learning-based GDM method. GDD is a deep neural network for spatial interpolation between sparsely distributed sensor measurements that was trained on an extensive data set of realistic-shaped synthetic gas plumes based on actual airflow measurements. As access to ground truth representations of gas distributions remains a challenge in GDM research, we make our data sets, along with our models, publicly available. We test and compare GDD with state-of-the-art models on synthetic and real-world data. Our findings demonstrate that GDD significantly outperforms existing models, demonstrating a 35% improvement in accuracy on synthetic data when measured using the Root Mean Squared Error over the entire distribution map. Notably, GDD appears to have superior capabilities in reconstructing the edges and characteristic shapes of gas plumes compared to traditional models. These potentials offer new possibilities for more accurate and efficient environmental monitoring, and we hope to inspire other researchers to explore learning-based GDM.
For mobile robots that operate in complex, uncontrolled environments, estimating air flow models can be of great importance. Aerial robots use air flow models to plan optimal navigation paths and to avoid turbulence-ridden areas. Search and rescue platforms use air flow models to infer the location of gas leaks. Environmental monitoring robots enrich pollution distribution maps by integrating the information conveyed by an air flow model. In this paper, we present an air flow modelling algorithm that uses wind data collected at a sparse number of locations to estimate joint probability distributions over wind speed and direction at given query locations. The algorithm uses a novel extrapolation approach that models the air flow as a linear combination of laminar and turbulent components. We evaluated the prediction capabilities of our algorithm with data collected with an aerial robot during several exploration runs. The results show that our algorithm has a high degree of stability with respect to parameter selection while outperforming conventional extrapolation approaches. In addition, we applied our proposed approach in an industrial application, where the characterization of a ventilation system is supported by a ground mobile robot. We compared multiple air flow maps recorded over several months by estimating stability maps using the Kullback-Leibler divergence between the distributions. The results show that, despite local differences, similar air flow patterns prevail over time. Moreover, we corroborated the validity of our results with knowledge from human experts.
For mobile robots that operate in complex, uncontrolled environments, estimating air flow models can be of great importance. Aerial robots use air flow models to plan optimal navigation paths and to avoid turbulence-ridden areas. Search and rescue platforms use air flow models to infer the location of gas leaks. Environmental monitoring robots enrich pollution distribution maps by integrating the information conveyed by an air flow model. In this paper, we present an air flow modelling algorithm that uses wind data collected at a sparse number of locations to estimate joint probability distributions over wind speed and direction at given query locations. The algorithm uses a novel extrapolation approach that models the air flow as a linear combination of laminar and turbulent components. We evaluated the prediction capabilities of our algorithm with data collected with an aerial robot during several exploration runs. The results show that our algorithm has a high degree of stability with respect to parameter selection while outperforming conventional extrapolation approaches. In addition, we applied our proposed approach in an industrial application, where the characterization of a ventilation system is supported by a ground mobile robot. We compared multiple air flow maps recorded over several months by estimating stability maps using the Kullback-Leibler divergence between the distributions. The results show that, despite local differences, similar air flow patterns prevail over time. Moreover, we corroborated the validity of our results with knowledge from human experts.
The objective of this Ph.D. thesis is the development and validation of a VTOL-based (Vertical Take Off and Landing) micro-drone for the measurement of gas concentrations, to locate gas emission sources, and to build gas distribution maps. Gas distribution mapping and localization of a static gas source are complex tasks due to the turbulent nature of gas transport under natural conditions [1] and becomes even more challenging when airborne. This is especially so, when using a VTOL-based micro-drone that induces disturbances through its rotors, which heavily affects gas distribution. Besides the adaptation of a micro-drone for gas concentration measurements, a novel method for the determination of the wind vector in real-time is presented. The on-board sensors for the flight control of the micro-drone provide a basis for the wind vector calculation. Furthermore, robot operating software for controlling the micro-drone autonomously is developed and used to validate the algorithms developed within this Ph.D. thesis in simulations and real-world experiments. Three biologically inspired algorithms for locating gas sources are adapted and developed for use with the micro-drone: the surge-cast algorithm (a variant of the silkworm moth algorithm) [2], the zigzag / dung beetle algorithm [3], and a newly developed algorithm called “pseudo gradient algorithm”. The latter extracts from two spatially separated measuring positions the information necessary (concentration gradient and mean wind direction) to follow a gas plume to its emission source. The performance of the algorithms is evaluated in simulations and real-world experiments. The distance overhead and the gas source localization success rate are used as main performance criteria for comparing the algorithms. Next, a new method for gas source localization (GSL) based on a particle filter (PF) is presented. Each particle represents a weighted hypothesis of the gas source position. As a first step, the PF-based GSL algorithm uses gas and wind measurements to reason about the trajectory of a gas patch since it was released by the gas source until it reaches the measurement position of the micro-drone. Because of the chaotic nature of wind, an uncertainty about the wind direction has to be considered in the reconstruction process, which extends this trajectory to a patch path envelope (PPE). In general, the PPE describes the envelope of an area which the gas patch has passed with high probability. Then, the weights of the particles are updated based on the PPE. Given a uniform wind field over the search space and a single gas source, the reconstruction of multiple trajectories at different measurement locations using sufficient gas and wind measurements can lead to an accurate estimate of the gas source location, whose distance to the true source location is used as the main performance criterion. Simulations and real-world experiments are used to validate the proposed method. The aspect of environmental monitoring with a micro-drone is also discussed. Two different sampling approaches are suggested in order to address this problem. One method is the use of a predefined sweeping trajectory to explore the target area with the micro-drone in real-world gas distribution mapping experiments. As an alternative sampling approach an adaptive strategy is presented, which suggests next sampling points based on an artificial potential field to direct the micro-drone towards areas of high predictive mean and high predictive variance, while maximizing the coverage area. The purpose of the sensor planning component is to reduce the time that is necessary to converge to the final gas distribution model or to reliably identify important parameters of the distribution such as areas of high concentration. It is demonstrated that gas distribution models can provide an accurate estimate of the location of stationary gas sources. These strategies have been successfully tested in a variety of real-world experiments in different scenarios of gas release using different gas sensors to verify the reproducibility of the experiments. The adaptive strategy was also successfully validated in simulations using predefined sweeping trajectories as reference criteria. The results of this Ph.D. thesis reflect the applicability of gas-sensitive microdrones in a variety of scenarios of gas release. Effective counteractive measures can be set in motion after accidents involving gas emissions with the aid of spatially resolved gas concentration and wind data collected with micro-drones. Monitoring of geochemically active regions, landfills, CO2 storage facilities, and the localization of gas leaks are further areas of application.
BAM Federal Institute for Materials Research and Testing, in cooperation with the AirRobot GmbH & Co. KG company, has developed a flying remote-controlled measuring system. The system is capable of operating in a variety of scenarios of gas emissions, e.g. exhaust gas from chimneys, flue gas in a fire, gas emissions in the case of an accident of chemical or hazardous goods or in the case of a terrorist act involving toxic gases. Thus it can measure the gas concentration in the immediate vicinity of the object which causes the emission. A further stage of extension is to enhance the system for plume tracking and identification of sources of hazardous gases.