Rapid Localization and Mapping Method Based on Adaptive Particle Filters.

With the development of autonomous vehicles, localization and mapping technologies have become crucial to equip the vehicle with the appropriate knowledge for its operation. In this paper, we extend our previous work by prepossessing a localization and mapping architecture for autonomous vehicles that do not rely on GPS, particularly in environments such as tunnels, under bridges, urban canyons, and dense tree canopies. The proposed approach is of two parts. Firstly, a K-means algorithm is employed to extract features from LiDAR scenes to create a local map of each scan. Then, we concatenate the local maps to create a global map of the environment and facilitate data association between frames. Secondly, the main localization task is performed by an adaptive particle filter that works in four steps: (a) generation of particles around an initial state (provided by the GPS); (b) updating the particle positions by providing the motion (translation and rotation) of the vehicle using an inertial measurement device; (c) selection of the best candidate particles by observing at each timestamp the match rate (also called particle weight) of the local map (with the real-time distances to the objects) and the distances of the particles to the corresponding chunks of the global map; (d) averaging the selected particles to derive the estimated position, and, finally, using a resampling method on the particles to ensure the reliability of the position estimation. The performance of the newly proposed technique is investigated on different sequences of the Kitti and Pandaset raw data with different environmental setups, weather conditions, and seasonal changes. The obtained results validate the performance of the proposed approach in terms of speed and representativeness of the feature extraction for real-time localization in comparison with other state-of-the-art methods.

Machine Learning Accelerated Discovery of Promising Thermal Energy Storage Materials with High Heat Capacity.

Thermal energy storage offers numerous benefits by reducing energy consumption and promoting the use of renewable energy sources. Thermal energy storage materials have been investigated for many decades with the aim of improving the overall efficiency of energy systems. However, finding solid materials that meet the requirement of high heat capacity has been a grand challenge for material scientists. Herewith, by training various machine learning models on 3377 high-quality data from full density functional theory (DFT) calculations, we efficiently search for potential materials with high heat capacity. We build four traditional machine learning models and two graph neural network models. Cross-comparison of the prediction performance and model accuracy was conducted among different models. The deeperGATGNN model exhibits high prediction accuracy and is used for predicting the heat capacity of 32,026 structures screened from the open quantum material database. We gain deep insight into the correlation between heat capacity and structure descriptors such as space group, prototype, lattice volume, atomic weight, etc. Twenty-two structures were predicted to possess high heat capacity, and the results were further validated with DFT calculations. We also identified one special structure, namely, MnIn2Se4, with space group no. 227 (Fd3̅m), that exhibits extremely high heat capacity, even higher than that of the Dulong-Petit limit at room temperature. This study paves the way for accelerating the discovery of novel thermal energy storage materials by combining machine learning with minimal DFT inquiry.

Data on one-dimensional vertical free swelling potential of soils and related soil properties.

Most of the damaging geo-hazards recorded in modern history are caused by soil swelling or expansion. Therefore, proper evaluation of a soil's capacity to swell is very crucial for the achievement of a secure and safe ground for civil infrastructures and related land developments which are founded on the soil. In order to simulate as well as estimate the heave that can occur under field conditions, laboratory one-dimensional oedometer vertical swell-strain testing are most frequently used. Hence, in this brief, one-dimensional swelling tests adopted to measure soil swelling on laboratory-engineered and natural soils covering various regions on the globe are reported. The testing standards and procedures followed in the measurement of one-dimensional swelling are those enumerated in the American Standards for Testing of Materials (ASTM) DDD698, and American Association of State Highways Transport Officials (AASHTO). Slight modifications to the measurement procedures (such as the use of different surcharge loading and custom-made consolidation rings) reflecting special laboratory testing conditions and for the purposes of comparisons, are also reported. Corresponding soil properties characterising the dataset includes moisture content, void ratio, specific gravity, unit weight, liquid limit, plastic limit, plasticity index, clay content, silt content, maximum dry unit weight, optimum moisture content, and soil activity index, all of which are known to bear either direct or indirect influences on soil. Determination of the state of compaction of the soils where applicable, are carried out based on the American Standards for Testing of Materials (ASTM) DDD698, Turkish Standards (TS), American Association of State Highways Transport Officials (AASHTO)and a combination of both standard and modified efforts. A total of 395 data samples on soil swelling potential are reported. With regards to the corresponding soil properties, a total of 219 data records of soil specific gravity, 321 data records of initial moisture content, 163 data records of void ratio, 273 data records of dry unit weight, 347 data records of liquid limit, 347 data records of plastic limit, 395 data records of plasticity index, 209 data records of activity index, 339 data records of clay content, 174 data records of silt content, 246 data records of optimum moisture content, 228 data records of maximum dry density and 347 data records of Unified Soil Classification System (USCS) are presented. Finally, the dataset of one-dimensional soil swelling described herein are intended to aid geotechnical engineers and researchers who are involved in statistical correlation studies, data analytics, and machine learning predictions using soft computing methods mostly aimed at evaluating soil expansion especially during the preliminary phases of soil investigation and foundation design.

IO-VNBD: Inertial and Odometry benchmark dataset for ground vehicle positioning.

Low-cost Inertial Navigation Sensors (INS) can be exploited for a reliable solution for tracking autonomous vehicles in the absence of GPS signals. However, position errors grow exponentially over time due to noises in the sensor measurements. The lack of a public and robust benchmark dataset has however hindered the advancement in the research, comparison and adoption of recent machine learning techniques such as deep learning techniques to learn the error in the INS for a more accurate positioning of the vehicle. In order to facilitate the benchmarking, fast development and evaluation of positioning algorithms, we therefore present the first of its kind large-scale and information-rich inertial and odometry focused public dataset called IO-VNBD (Inertial Odometry Vehicle Navigation Benchmark Dataset). The vehicle tracking dataset was recorded using a research vehicle equipped with ego-motion sensors on public roads in the United Kingdom, Nigeria, and France. The sensors include a GPS receiver, inertial navigation sensors, wheel-speed sensors amongst other sensors found in the car, as well as the inertial navigation sensors and GPS receiver in an Android smart phone sampling at 10 Hz. A diverse number of driving scenarios were captured such as traffic congestion, round-abouts, hard-braking, etc. on different road types (e.g. country roads, motorways, etc.) and with varying driving patterns. The dataset consists of a total driving time of about 40 h over 1,300 km for the vehicle extracted data and about 58 h over 4,400 km for the smartphone recorded data. We hope that this dataset will prove valuable in furthering research on the correlation between vehicle dynamics and dependable positioning estimation based on vehicle ego-motion sensors, as well as other related studies.