-
Vision-based navigation and obstacle avoidance via deep reinforcement learning
Authors:
Paul Blum,
Peter Crowley,
George Lykotrafitis
Abstract:
Development of navigation algorithms is essential for the successful deployment of robots in rapidly changing hazardous environments for which prior knowledge of configuration is often limited or unavailable. Use of traditional path-planning algorithms, which are based on localization and require detailed obstacle maps with goal locations, is not possible. In this regard, vision-based algorithms h…
▽ More
Development of navigation algorithms is essential for the successful deployment of robots in rapidly changing hazardous environments for which prior knowledge of configuration is often limited or unavailable. Use of traditional path-planning algorithms, which are based on localization and require detailed obstacle maps with goal locations, is not possible. In this regard, vision-based algorithms hold great promise, as visual information can be readily acquired by a robot's onboard sensors and provides a much richer source of information from which deep neural networks can extract complex patterns. Deep reinforcement learning has been used to achieve vision-based robot navigation. However, the efficacy of these algorithms in environments with dynamic obstacles and high variation in the configuration space has not been thoroughly investigated. In this paper, we employ a deep Dyna-Q learning algorithm for room evacuation and obstacle avoidance in partially observable environments based on low-resolution raw image data from an onboard camera. We explore the performance of a robotic agent in environments containing no obstacles, convex obstacles, and concave obstacles, both static and dynamic. Obstacles and the exit are initialized in random positions at the start of each episode of reinforcement learning. Overall, we show that our algorithm and training approach can generalize learning for collision-free evacuation of environments with complex obstacle configurations. It is evident that the agent can navigate to a goal location while avoiding multiple static and dynamic obstacles, and can escape from a concave obstacle while searching for and navigating to the exit.
△ Less
Submitted 9 November, 2022;
originally announced November 2022.
-
Quantification of fracture roughness by change probabilities and Hurst exponents
Authors:
Tim Gutjahr,
Sina Hale,
Karsten Keller,
Philipp Blum,
Steffen Winter
Abstract:
The objective of the current study is to utilize an innovative method called 'change probabilities' for describing fracture roughness. In order to detect and visualize anisotropy of rock joint surfaces, the roughness of one-dimensional profiles taken in different directions is quantified. The central quantifiers, 'change probabilities', are based on counting monotone changes in discretizations of…
▽ More
The objective of the current study is to utilize an innovative method called 'change probabilities' for describing fracture roughness. In order to detect and visualize anisotropy of rock joint surfaces, the roughness of one-dimensional profiles taken in different directions is quantified. The central quantifiers, 'change probabilities', are based on counting monotone changes in discretizations of a profile. These probabilities, which are usually varying with the scale, can be reinterpreted as scale-dependent Hurst exponents. For a large class of Gaussian stochastic processes change probabilities are shown to be directly related to the classical Hurst exponent, which generalizes a relationship known for fractional Brownian motion. While being related to this classical roughness measure, the proposed method is more generally applicable, increasing therefore the flexibility of modeling and investigating surface profiles. In particular, it allows a quick and efficient visualization and detection of roughness anisotropy and scale dependence of roughness.
△ Less
Submitted 16 November, 2021;
originally announced November 2021.
-
MobilityCoins -- A new currency for the multimodal urban transportation system
Authors:
Klaus Bogenberger,
Philipp Blum,
Florian Dandl,
Lisa-Sophie Hamm,
Allister Loder,
Patrick Malcolm,
Martin Margreiter,
Natalie Sautter
Abstract:
The MobilityCoin is a new, all-encompassing currency for the management of the multimodal urban transportation system. MobilityCoins includes and replaces various existing transport policy instruments while also incentivizing a shift to more sustainable modes as well as empowering the public to vote for infrastructure measures.
The MobilityCoin is a new, all-encompassing currency for the management of the multimodal urban transportation system. MobilityCoins includes and replaces various existing transport policy instruments while also incentivizing a shift to more sustainable modes as well as empowering the public to vote for infrastructure measures.
△ Less
Submitted 29 July, 2021; v1 submitted 28 July, 2021;
originally announced July 2021.
-
Why Monday never wins -- An Example of the Secretary Problem
Authors:
Peter Blum,
Marc Wenskat
Abstract:
We analyzed the winning statistics of the German TV show 'Das Perfekte Dinner', a competition where the contestants rate each other. We did a comparison of the original and the celebrity version of the show and also implemented a rescaling of the points to reduce the influences of subjective grading. We found that there is a strong dependency of the winning rate on the position of performing, not…
▽ More
We analyzed the winning statistics of the German TV show 'Das Perfekte Dinner', a competition where the contestants rate each other. We did a comparison of the original and the celebrity version of the show and also implemented a rescaling of the points to reduce the influences of subjective grading. We found that there is a strong dependency of the winning rate on the position of performing, not only for the first, but for all candidates. Furthermore, we concluded that there is a positive bias in giving points, after the contestants have already cooked. The comparison with a simpler optimizing strategy, the so called secretary problem, showed a lot of similarities.
△ Less
Submitted 4 June, 2020;
originally announced June 2020.
-
What is the difference? Blazhko and non-Blazhko RRab stars and the special case of V123 in M3
Authors:
J. Jurcsik,
P. Smitola,
G. Hajdu,
C. Pilachowski,
K. Kolenberg,
Á. Sódor,
G. Fűrész,
A. Moór,
E. Kun,
A. Saha,
P. Prakash,
P. Blum,
I. Tóth
Abstract:
In an extended photometric campaign of RR Lyrae variables of the globular cluster M3, an aberrant light-curve, non-Blazhko RRab star, V123, was detected. Based on its brightness, colors and radial velocity curve, V123 is a bona fide member of M3. The light curve of V123 exhibits neither a bump preceding light minimum, nor a hump on the rising branch, and has a longer than normal rise time, with a…
▽ More
In an extended photometric campaign of RR Lyrae variables of the globular cluster M3, an aberrant light-curve, non-Blazhko RRab star, V123, was detected. Based on its brightness, colors and radial velocity curve, V123 is a bona fide member of M3. The light curve of V123 exhibits neither a bump preceding light minimum, nor a hump on the rising branch, and has a longer than normal rise time, with a convex shape. Similar shape characterizes the mean light curves of some large-modulation-amplitude Blazhko stars, but none of the regular RRab variables with similar pulsation periods. This peculiar object thus mimics Blazhko variables without showing any evidence of periodic amplitude and/or phase modulation. We cannot find any fully convincing answer to the peculiar behavior of V123, however, the phenomenon raises again the possibility that rotation and aspect angle might play a role in the explanation of the Blazhko phenomenon, and some source of inhomogeneity acts (magnetic field, chemical inhomogeneity) that deforms the radial pulsation of Blazhko stars during the modulation.
△ Less
Submitted 20 September, 2013;
originally announced September 2013.
-
A New Natural Gamma Radiation Measurement System for Marine Sediment and Rock Analysis
Authors:
M. A. Vasiliev,
P. Blum,
G. Chubarian,
R. Olsen,
C. Bennight,
T. Cobine,
D. Fackler,
M. Hastedt,
D. Houpt,
Z. Mateo,
Y. B. Vasilieva
Abstract:
A new high-efficiency and low-background system for the measurement of natural gamma radioactivity in marine sediment and rock cores retrieved from beneath the seabed was designed, built, and installed on the JOIDES Resolution research vessel. The system includes eight large NaI(Tl) detectors that measure adjacent intervals of the core simultaneously, maximizing counting times and minimizing stati…
▽ More
A new high-efficiency and low-background system for the measurement of natural gamma radioactivity in marine sediment and rock cores retrieved from beneath the seabed was designed, built, and installed on the JOIDES Resolution research vessel. The system includes eight large NaI(Tl) detectors that measure adjacent intervals of the core simultaneously, maximizing counting times and minimizing statistical error for the limited measurement times available during drilling expeditions. Effect to background ratio is maximized with passive lead shielding, including both ordinary and low-activity lead. Large-area plastic scintillator active shielding filters background associated with the high-energy part of cosmic radiation. The new system has at least an order of magnitude higher statistical reliability and significantly enhances data quality compared to other offshore natural gamma radiation (NGR) systems designed to measure geological core samples. Reliable correlations and interpretations of cored intervals are possible at rates of a few counts per second.
△ Less
Submitted 11 November, 2010;
originally announced November 2010.
-
Construction and Performance of a Micro-Pattern Stereo Detector with Two Gas Electron Multipliers
Authors:
T. Barvich,
P. Bluem,
M. Erdmann,
M. Fahrer,
K. Kaercher,
F. Kuehn,
D. Moermann,
Th. Mueller,
D. Neuberger,
F. Roederer,
H. J. Simonis,
A. Skiba,
W. H. Thuemmel,
Th. Weiler,
S. Weseler
Abstract:
The construction of a micro-pattern gas detector of dimensions 40x10 cm**2 is described. Two gas electron multiplier foils (GEM) provide the internal amplification stages. A two-layer readout structure was used, manufactured in the same technology as the GEM foils. The strips of each layer cross at an effective crossing angle of 6.7 degrees and have a 406 um pitch. The performance of the detecto…
▽ More
The construction of a micro-pattern gas detector of dimensions 40x10 cm**2 is described. Two gas electron multiplier foils (GEM) provide the internal amplification stages. A two-layer readout structure was used, manufactured in the same technology as the GEM foils. The strips of each layer cross at an effective crossing angle of 6.7 degrees and have a 406 um pitch. The performance of the detector has been evaluated in a muon beam at CERN using a silicon telescope as reference system. The position resolutions of two orthogonal coordinates are measured to be 50 um and 1 mm, respectively. The muon detection efficiency for two-dimensional space points reaches 96%.
△ Less
Submitted 4 October, 2001;
originally announced October 2001.
