Achim J. Lilienthal - Academia.edu (original) (raw)
Achim J. Lilienthal is professor of Computer Science and head of the Mobile Robotics and Olfaction (MRO) Lab at Örebro University. His core research interests are in perception systems in unconstrained, dynamic environments. Typically based on approaches that leverage domain knowledge and Artificial Intelligence his research work addresses mobile robot olfaction, rich 3D perception, navigation of autonomous transport robots, human robot interaction and mathematics education research. Achim J. Lilienthal obtained his Ph.D. in computer science from Tübingen University. The Ph.D. thesis addresses gas distribution mapping and gas source localisation with mobile robots. He is author/coauthor of more than 250 refereed conference papers and journal articles.
less
Uploads
Papers by Achim J. Lilienthal
This paper addresses the problem of localising a static gas source in an uncontrolled indoor envi... more This paper addresses the problem of localising a static gas source in an uncontrolled indoor environment by a mobile robot. In contrast to previous works, especially the condition of an environment that is not artificially ventilated to pro- duce a strong unidirectional airflow is considered. Here, the propagation of the analyte molecules is dominated by turbulence and convection flow rather
New approach for automatically building symbolic relational descriptions of static configurations... more New approach for automatically building symbolic relational descriptions of static configurations of objects to be manipulated. • Geometrical reasoning to extract ACT relations that indicate which objects acts upon which other object(s). • Static equilibrium analysis to evaluate support relations between objects. • A graph of objects as vertices and relations as edges represents the extracted symbolic relations. Actual scene (top left), point cloud (top right), model after recognition of the box objects (bottom left) and resulting relational scene representation (bottom right).
The author has requested enhancement of the downloaded file. All in-text references underlined in... more The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate. The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.
Due to its environmental, economical and safety implications, methane leak detection is a crucial... more Due to its environmental, economical and safety implications, methane leak detection is a crucial task to address in the biogas production industry. In this paper, we introduce Gasbot, a robotic platform that aims to automatize methane emission monitoring in landfills and biogas production sites. The distinctive characteristic of the Gasbot platform is the use of a Tunable Laser Absorption Spectroscopy (TDLAS) sensor, along with a novel gas distribution algorithm to generate methane concentration maps of indoor and outdoor exploration areas. The Gasbot platform has been tested in two different scenarios: an underground corridor, where a pipeline leak was simulated and in a decommissioned landfill site, where an artificial methane emission source was introduced.
Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from ... more Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection # 16 October 7 th , 2012 3. A proper selection of the kernel bandwidth determines the performance of the predictive model o For GDM, the performance of a model is measured using the NLPD o Captures both properties: mean and variance predictions Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection Operations required to update and evaluate the models K x m x N x N g # 17 October 7 th , 2012 3. A proper selection of the kernel bandwidth determines the performance of the predictive model o For GDM, the performance of a model is measured using the NLPD o Captures both properties: mean and variance predictions Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection Operations required to update and evaluate the models K x m x N x N g
This paper addresses the problem of localising a static gas source in an uncontrolled indoor envi... more This paper addresses the problem of localising a static gas source in an uncontrolled indoor environment by a mobile robot. In contrast to previous works, especially the condition of an environment that is not artificially ventilated to pro- duce a strong unidirectional airflow is considered. Here, the propagation of the analyte molecules is dominated by turbulence and convection flow rather
New approach for automatically building symbolic relational descriptions of static configurations... more New approach for automatically building symbolic relational descriptions of static configurations of objects to be manipulated. • Geometrical reasoning to extract ACT relations that indicate which objects acts upon which other object(s). • Static equilibrium analysis to evaluate support relations between objects. • A graph of objects as vertices and relations as edges represents the extracted symbolic relations. Actual scene (top left), point cloud (top right), model after recognition of the box objects (bottom left) and resulting relational scene representation (bottom right).
The author has requested enhancement of the downloaded file. All in-text references underlined in... more The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate. The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.
Due to its environmental, economical and safety implications, methane leak detection is a crucial... more Due to its environmental, economical and safety implications, methane leak detection is a crucial task to address in the biogas production industry. In this paper, we introduce Gasbot, a robotic platform that aims to automatize methane emission monitoring in landfills and biogas production sites. The distinctive characteristic of the Gasbot platform is the use of a Tunable Laser Absorption Spectroscopy (TDLAS) sensor, along with a novel gas distribution algorithm to generate methane concentration maps of indoor and outdoor exploration areas. The Gasbot platform has been tested in two different scenarios: an underground corridor, where a pipeline leak was simulated and in a decommissioned landfill site, where an artificial methane emission source was introduced.
Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from ... more Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection # 16 October 7 th , 2012 3. A proper selection of the kernel bandwidth determines the performance of the predictive model o For GDM, the performance of a model is measured using the NLPD o Captures both properties: mean and variance predictions Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection Operations required to update and evaluate the models K x m x N x N g # 17 October 7 th , 2012 3. A proper selection of the kernel bandwidth determines the performance of the predictive model o For GDM, the performance of a model is measured using the NLPD o Captures both properties: mean and variance predictions Cross Validation (CV) is the most commonly used method to select an optimal bandwidth o from a [ 1 , 2 ,…, m ] This requires to train and test K x m models CV Bandwidth Selection Operations required to update and evaluate the models K x m x N x N g