TalTech Data Repository
Not a member yet
221 research outputs found
Sort by
Synthetic dataset for warehouse equipment and pallet recognition (Videos)
No full text<p>This dataset provides a collection of synthetic images simulating various warehouse scenarios, specifically designed for training and evaluating computer vision models for equipment and pallet recognition tasks. The images feature a diverse range of warehouse environments, including different lighting conditions, object arrangements, and occlusions.</p><p>This dataset was created as part of a project funded by <a href="https://www.etis.ee/Portal/Projects/Display/196065ff-51bc-48c8-9790-62115a68dec3"><strong>AI & ROBOTICS ESTONIA (EDIH)</strong></a></p><p>Due to upload size limits the dataset was split into 4 pieces.</p><ul><li><a href="/records/nmw7y-41a87"><strong>Camera 1</strong></a></li><li><a href="/records/3ytc7-2z453"><strong>Camera 2</strong></a></li><li><a href="/records/c6182-bgd05"><strong>Models</strong></a></li><li><strong>Videos</strong></li></ul><p>For an in depth explaination about the dataset creation and model training process see the thesis in <i>Related works</i>.</p><p><strong>Scene description:</strong><br>Forklift carries a pallet between warehouse racks. It deposits the pallet on a shelf.<br>Another pallet is then picked up from the floor and carried away.</p><p>All cameras capture the same scene from different angles. They are all syncronized to allow for multi camera tracking.</p><p><i><strong>Note: </strong></i>Camera1.mp4 and Camera2.mp4 do not have any relations to datasets Camera 1 and 2.</p>
Enhancing Digital Permitting through Improved Data Quality and Standardization
No full text<p>The global construction sector faces data fragmentation, lack of interoperability, and inefficient workflows, complicating building permitting and as such delaying projects. This research focuses on how digital transformation can enhance data interoperability, consistency, and quality using ontologies and Semantic Web technologies. Initial analysis suggests these technologies can streamline code compliance checking and improve building permit workflows. By using the Estonian Building Registry as a case, this research demonstrates that the public sector can drive digitalization in construction. It highlights the critical role of high quality building data in enhancing building permit processes and addresses the current lack of a comprehensive data architecture. By adding a semantic layer to the e-construction platform we can address the data fragmentation and interoperability issues. </p>
<p> </p>
Simulated gait dataset for abnormality detection
No full text<p>The dataset is created by collecting volunteers walking data using inertial motion units (IMU) placed on the forefoot. Data collection device was Shimmer S3 IMU.</p><p>Volunteers simulated eight different common gait deviations focusing on lower extremities.</p><p>Simulation was recreating actual patients' gait patterns, under the instructions of professional physiotherapist.</p><p>Each gait recording contains normal ("ok") and abnormal ("ab") step patterns, as well as turn ("turn") steps.</p><p>Start of the recording is shown by label "start" and "stop" of the recording is shown by label stop.</p><p>In total, there are 22 different persons, who simulated different combinations of abnormal step patterns. </p><p>Data was annotated by the definitive shapes of step patterns using gyroscope magnitude and gait recording videos for reference.</p><p> </p><p>In the main folder of "depersonalyzed_data" are folders containing data for each person separately.</p><p>Then for each person folder, there are folders named after gait types, which were simulated.</p><p>In each gait type folder, there are one to several gait recordings in CSV files, and corresponding labels in the txt files.</p><p>The delimiter in the CSV file is tab. In the TXT file, each label is on the new line.</p><p> </p><p>CSV file has the following structure. Columns: time in ms, accelerometer axis x,y,z in m/(s^2) and gyroscope axis x,y,z in deg/s. Rows: first row is containing standard Shimmer values, which are unique for each column.</p><p>Second row contains units of measurement, starting from third row is data itself, where decimal is separated by full stop.</p><p>Sampling rate is set to 256Hz. Accelerometer range is 8G and gyroscope range is 1000 deg/s.</p><p> </p><p>TXT file has the following structure. Columns: start of the label, end of label and value of the label.</p><p> </p><p>Data has been used in the publications listed below, first for classification and then for real-time in-step anomaly detection.</p><p>The studies approved by Estonian Research Ethics Committee of the National Institute for Health Development, permission No.818. The participants provided their written informed consent to participate in this study.</p><p> </p><p>To access the data please use link below:</p><p>https://drive.google.com/drive/folders/10FDCG4xpRwhuYfh9eatbPI0D0JsHxOe_?usp=sharing</p>
Currents and CTD measurements data in the vicinity of Keri Island in 2018
No full text<p><strong>Data description of currents and temperature and salinity profiles in the vicinity of Keri Island in the Gulf of Finland in the Baltic Sea in summer/autumn 2018.</strong></p><p>Oceanographic measurements, encompassing currents, water temperature, and salinity, were conducted at two proximate locations near Keri Island. The bottom-mounted ADCP (Acoustic Doppler Current Profiler, Workhorse Sentinel, Teledyne RDI, 300 kHz) measured current profiles for two months, from August to September 2018, at the KeriN measurement station, approximately 2 km north of Keri Island. Simultaneously, temperature and salinity profiles were recorded for nearly one month by a bottom-mounted autonomous CTD profiler (Flydog Solutions) at the KeriN from 26 July 2018 to 17 August 2018. CTD (Conductivity, Temperature, Depth) profiles were recorded from 3 to 98 meters at 3-hour intervals and ADCP data from 8 to 106 meters at 1-hour intervals. The horizontal separation between the locations of CTD and ADCP profilers was 60 m.</p><p>The bottom-mounted RDCP (Recording Doppler Current Meter, Aanderaa Data Instruments AS) measured current profiles for one month, spanning 28 August to 2 October 2018, at the KeriS measurement station, approximately 1.1 km east of the island. Simultaneously, temperature and salinity profiles were obtained for two months by a buoy-mounted automatic CTD profiler (Idronaut) at the KeriS from 28 August 2018 to 5 October 2018. CTD data, available at 6-hour intervals, covered depths from 2 to 44 meters, and RDCP data spanned depths from 6.8 to 42 meters at 1-hour intervals. The separation between the locations of CTD and RDCP profilers was 300 m.</p><p>High-resolution wind data for the measurement period were obtained from the Tallinnamadal Lighthouse, about 20 km west of the measurement site. Wind speed and direction sensors (Aanderaa) at a height of 31 m recorded data every 5 minutes. </p>
Simulation data used in management CCL for RAN slice performance improvement by subslicing
No full text<p>Simulation data used in management CCL for RAN slice performance improvement by subslicing</p><p>------------------------------------------------------------------------------------------</p><p>Data used in PhD thesis and publication 4:</p><p>Folder NNtrainingdata\ contains training data used to train NN to decide subslice merge, split or no change in MCCL.</p><p>Folder Results_init\ contains slice simulation results of initialization of slice configuration if used subslice split algorithms in MCCL.</p><p>Folder Results_runtime_paper4\ contains slice simulation results of runtime if used subslice split algorithms in MCCL. Results are used in publication 4 of thesis.</p><p>This work has received funding partly from the European Union's Horizon 2020 Research and Innovation Program under Grant 951867 '5G-ROUTES' and Grant 101058505 '5G-TIMBER'. This work in the project "ICT programme" was supported by the European Union through European Social Fund, and TAR16013 Center of Excellence 'EXCITE IT'.</p>
Dataset of Forklift-Deployed Multi-Sensor Setup for Indirect Tracking of Markerless Industrial Products
No full text<p>Dataset of the time-synchronized multisensor data, collected by a total of four sensors, deployed on the full-scale forklift, operating in the industrial environment. The dataset contains the experimentally collected data for the Indirect Tracking project, as well as its selective processing results. The following initial data is provided: Positioning data of the forklift, collected by using <i><strong>Eliko UWB RTLS system</strong></i>; Selective inertial data of the forklift heading (Z-axis angular velocity), collected by the gyroscope of the <i><strong>IMU unit Bosh bno055</strong></i>; Fork elevation data, collected by <i><strong>Miran MPS series draw wire encoder</strong></i>; Distance to the object in front of the forklift data, collected by <i><strong>ultrasonic distance sensor SEN0208</strong></i>.</p><p>The dataset also contains certain results of the initial data processing, used in the context of the Indirect Tracking project. The dataset is provided in the Indirect_Tracking_Dataset.csv file, and contains the following data sections: </p><p><strong>FORKLIFT POSITIONING DATA </strong>includes the initial forklift positioning data, including real-time 2D coordinates of the forklift: <strong>X</strong> [m] and <strong>Y</strong> [m], collected by the UWB RTLS system at the update rate of 5Hz, along with the corresponding time within the experimental campaign <strong>TIME.xy</strong> [s] and delta time of the <strong>dt.xy </strong>[s] between positioning data samples.</p><p><strong>FORKLIFT IMU INPUT </strong>includes the angular velocity of the forklift heading <strong>GYR_Z</strong> [deg/s], collected by the onboard gyroscope at 100Hz update rate, along with the corresponding time within the test campaign <strong>TIME.xy</strong> [s] and delta time of the <strong>dt.xy </strong>[s] between the data samples.</p><p><strong>FORKLIFT HEADING DATA</strong> includes the calculated forklift heading data <strong>ATKF.YAW</strong> [deg], based on the above-mentioned inertial and positioning data, by using the ATKF (Adaptive Tandem Kalman Filter) heading estimation algorithm. This section also includes the estimation of the expected true forklift heading <strong>TRU.YAW</strong> [deg].</p><p><strong>ESTIMATED FORKLIFT FORK COORDINATES & FORK SENSORS' DATA </strong>includes the collected and calculated data, related to the fork (tynes) of the used forklift. The provided 2D coordinates of the fork area <strong>FORK.X</strong> [m] and <strong>FORK.Y</strong> [m] are calculated by using the above-mentioned forklift positioning and (ATKF estimated) heading data. This section also provides the initial measurement data of the fork <strong>elevation sensor </strong>(<strong>FORK.Z</strong> [m]), reflecting the Z coordinate of the fork, and the distance data collected by the ultrasonic <strong>Distance sensor</strong> [m], reflecting the occupancy of the fork area. Fork elevation and distance sensors' data were simultaneously collected by the onboard MCU with the appropriate update rate of 12Hz).</p><p><strong>INTERACTION EVENT </strong>reflects the sequential number and the exact period of the forklift interaction with the payload (pick-up or drop-down event).</p><p>As in the tested indirect tracking method, the location of the payload pick-up or drop-down is formed over time from 2D (X&Y) and elevation (Z) coordinates, the <strong>INDIR.COORD.SAVED</strong> reflects the particular event coordinates part (X&Y, or Z), saved at the exact moment. The exact moment of the event coordinates saving is calculated by using the A-PDD (Automatic Pick-up & Drop-down Detection) algorithm, based on the initial data of the aforementioned fork elevation and distance sensors.</p><p>The collected data reflects the experimental positioning of two industrial payloads. Sections <strong>STATUS & POSITION OF INDIRECTLY TRACKED PAYLOAD 1</strong> and <strong>STATUS & POSITION OF INDIRECTLY TRACKED PAYLOAD 2 </strong>separately provide the momentary status information <strong>PYLD1.STATUS</strong> & <strong>PYLD2.STATUS </strong>of each tracked payload. Each payload has one of four status values: <i>Stored</i> - payload is stored at the known coordinates; <i>Transported</i> - payload is being transported by the forklift and indirectly tracked in real-time; <i>Pick-up</i> - payload pick-up process (event) is ongoing and its coordinates are being defined; Drop-down - payload drop-down process (event) is ongoing and its coordinates are being defined. These sections also provide the corresponding indirectly tracked 3D coordinates of each payload (<strong>INDIR.PYLD1.X</strong> [m], <strong>INDIR.PYLD1.Y</strong> [m], <strong>INDIR.PYLD1.Z</strong> [m]) and (<strong>INDIR.PYLD2.X</strong> [m], <strong>INDIR.PYLD2.Y</strong> [m], <strong>INDIR.PYLD2.Z</strong> [m])</p><p><strong>UWB INITIAL PERFORMANCE in INDIRECT METHOD </strong>section reflects the momentary positioning quality of the used UWB RTLS system in the initial forklift tracking, and therefore, in the indirect payload localization. The section provides the momentary number of UWB anchor units in the line of sight with the forklift onboard UWB tag <strong>INDIR.ANCH_in_LOS </strong>(min. required 3), and the momentary status <strong>INDIR.POS.SUCCESS </strong>of successful coordinates calculation by the UWB system (<i>yes</i> or <i>no</i>)<strong>.</strong></p><p><strong>POSITION OF DIRECTLY TRACKED PAYLOAD 2 </strong>includes the real-time coordinates of payload 2 (<strong>DIR.PYLD2.X</strong> [m], <strong>DIR.PYLD2.Y </strong>[m], <strong>DIR.PYLD2.Z</strong> [m]), independently tracked by the directly attached UWB tag. </p><p><strong>UWB INITIAL PERFORMANCE in DIRECT METHOD </strong>section reflects the positioning quality of the used UWB RTLS system regarding the UWB tag, directly attached to the payload 2. The momentary number of UWB anchor units in the line of sight with the payload attached UWB tag <strong>DIR.ANCH_in_LOS </strong>(min. required 4 for 3D positioning) is provided in this section along with the momentary status <strong>DIR.POS.SUCCESS </strong>of successful coordinates calculation (<i>yes</i> or <i>no</i>).</p><p>Section <strong>TRUE COORDINATES OF INTERACTED PAYLOADS</strong> provides the manually measured true coordinates (<strong>PYLD.TRU.X</strong> [m], <strong>PYLD.TRU.Y</strong> [m], <strong>PYLD.TRU.Z</strong> [m]) of the interacted payload at the moment of the corresponding event.</p>
MarCyb dataset
No full text<p>Digitalisation is a global trend affecting many critical infrastructures, including maritime transportation. Incorporating information and operational technologies into transportation processes significantly enhances efficiency but also introduces substantial risks originating from cyberspace.</p><p>To support researchers and cyber experts, we share maritime navigation network data with and without cyber attacks.</p>
Dataset of Estonian phosphorite dissolution in hydrochloric acid
No full text<p>The file with the name experimental_measurements.csv contains experimental data for parallel experiments of Estonian phosphorite dissolution in hydrochloric acid and has columns of </p><ul><li>time - seconds from the beginning of the experiment</li><li> M - acid molarity,</li><li>size - phosphorite particle size</li><li> F_con, F_pro, P_con, P_pro, Ca_con, Ca_pro, Fe_con, Fe_pro - compounds concentration and solubility pocentage. </li><li>Try - number of experiment</li><li>ph_filtered - measured pH, smoothed by noise-reduction filter</li></ul><p>The file with the name experimental_averaged.xlsx has averaged experimental data for each experiment setup. </p>