{"success":true,"data":{"author":"Colorado State University and Western Michigan University (Rabinowitz, Gaikwad, White, Bradley, Asher)","bucketAccessType":"public","bucketName":"AD-Assurance-Lab/connected-eco-driving-doe-ee0008468","bucketPrefix":"","category":"Robotics","cloudProvider":"huggingface","createdAt":"2026-11-21T22:15:32.964Z","description":"Real-world connected-vehicle data from urban drives in Fort Collins, Colorado, in 2019, collected for vehicle velocity prediction and predictive energy management under U.S. Department of Energy Vehicle Technologies Office project DE-EE0008468 (Colorado State University with NREL, the cities of Fort Collins and Denver, and Western Michigan University). An instrumented Chevrolet Malibu drove arterial loops with a Mobileye 6 camera, Bosch mid-range radars, a ZED stereo camera and a CAN/OBD logger; volunteer cars with Freematics GPS/IMU/OBD loggers stood in for vehicle-to-vehicle (V2V) messages, and the city supplied traffic-signal phase and timing (SPaT) and BlueTOAD segment travel times (vehicle-to-infrastructure, V2I). The speed-prediction/ folder holds ML-ready tables, one row per distance step, with ego speed and position, the next five signals' phase and distance, segment speeds and V2V speed statistics, loadable with Hugging Face datasets. raw/ holds Vector BLF ADAS logs with MATLAB/CSV decodes and logger CSVs; processed/ holds synchronized MATLAB structures; results/, maps/ and a ZED stereo SVO video are included. About 14 GB (measured 2026-11-21). CC BY 4.0, copyright Colorado State University and Western Michigan University; cite Rabinowitz et al., SAE Technical Paper 2020-01-0736.","displayTitle":"Connected Eco-Driving, Fort Collins 2019: Synchronized Vehicle, ADAS, V2V and Traffic-Signal (V2I) Data for Velocity Prediction (DOE EE0008468)","huggingFaceUrl":"https://huggingface.co/datasets/AD-Assurance-Lab/connected-eco-driving-doe-ee0008468","id":"71710184-b30f-5e77-b1e3-d570ef39c72e","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/AD-Assurance-Lab/connected-eco-driving-doe-ee0008468","objectCountEstimate":129,"probeStatus":"ok","referenceUrl":"https://doi.org/10.4271/2020-01-0736","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"AD-Assurance-Lab/connected-eco-driving-doe-ee0008468\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"AD-Assurance-Lab/connected-eco-driving-doe-ee0008468\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=AD-Assurance-Lab, license:cc-by-4.0, size_categories:100K<n<1M, format:parquet, downloads=452"],"format":"mixed","generatedAt":"2026-11-21T22:31:27.537Z","layout":{"example":"processed/nn-training/D01Cleaned.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"GPS Time, Device Time, Longitude, Latitude,Speed (GPS)(km/h),next10sec,Label, Horizontal Dilution of Precision, Altitude(m), Bearing, Gravity X(G), Gravity Y(G), Gravity Z(G),Absolute Throttle Position B(%),Acceleration Sensor(Total)(g),Acceleration Sensor(X axis)(g),Acceleration Sensor(Y axis)(g),Acceleration Sensor(Z axis)(g),Accelerator PedalPosition D(%),Accelerator PedalPosition E(%),Accelerator PedalPosition F(%),Actual engine % torque(%),Air Fuel Ratio(Commanded)(:1),Air Fuel Ratio(Measured)(:1),Ambi","encoding":"utf8","key":"processed/nn-training/D01Cleaned.csv","truncatedAt":512}],"sampleObjects":["NOTICE","SHA256SUMS","maps/FtCoMap.PNG","maps/FtCoMapWithTSPandV2I.png","processed/2019-05-15/ADAS_ASDT.mat"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\npretty_name: Connected eco-driving, Fort Collins 2019 (DOE DE-EE0008468)\ntask_categories:\n- time-series-forecasting\ntags:\n- vehicle-velocity-prediction\n- connected-vehicles\n- v2v\n- v2i\n- signal-phase-and-timing\n- adas\n- eco-driving\n- energy-management\nsize_categories:\n- 100K<n<1M\nconfigs:\n- config_name: speed-prediction\n  data_files:\n  - split: aug_2019\n    path: speed-prediction/2019-08-01.parquet\n  - split: oct_2019\n    path: speed-prediction/2019-10-21_22.parquet\n---\n\n# Connected eco-driving, Fort Collins 2019\n\nReal-world, synchronized vehicle, ADAS, vehicle-to-vehicle (V2V) and infrastructure (V2I)\ndata from urban drives in Fort Collins, Colorado, in 2019, collected for **vehicle velocity\nprediction** and predictive energy management. Part of the U.S. Department of Energy Vehicle\nTechnologies Office project DE-EE0008468, \"Mobility and Energy Improvements for Fort Collins\nand Denver\" (Colorado State University with NREL, the cities of Fort Collins and Denver, and\nWestern Michigan University). Described in:\n\n> Rabinowitz, Gaikwad, White, Bradley, Asher. *Synchronous and Open, Real World, Vehicle,\n> ADAS, and Infrastructure Data Streams for Automotive Machine Learning Algorithms Research.*\n> SAE Technical Paper 2020-01-0736, 2020. https://doi.org/10.4271/2020-01-0736\n\nCode, documents and the full publication list:\n[AD-Assurance-Lab/connected-eco-driving--doe-ee0008468--main](https://github.com/AD-Assurance-Lab/connected-eco-driving--doe-ee0008468--main).\n\n## How it was collected\n\nAn instrumented Chevrolet Malibu (the ego vehicle) drove loops on arterials around central\nFort Collins (College, Prospect, Shields, Mulberry, Drake, Lemay) with a Mobileye 6 camera,\nBosch mid-range radars, a ZED stereo camera and a CAN/OBD logger. Volunteers drove their own\ncars nearby with Freematics GPS/IMU/OBD loggers to stand in for V2V messages. The City of\nFort Collins supplied traffic-signal phase and timing and BlueTOAD segment travel times.\n\n| Day | Folder | Ego | V2V drivers | Infrastructure |\n|---|---|---|---|---|\n| 2019-05-15 | `raw/2019-05-15/` | ADAS log, ZED video | D01-D04 | signal timing, segment speeds |\n| 2019-07-03 | `raw/2019-07-03/` | logger, ADAS log | D02, D03, D05, D06 | signal timing, BlueTOAD segments |\n| 2019-08-01 | `raw/2019-08-01/` | logger, ADAS log | D02, D03 | signal timing, segment speeds |\n| 2019-10-21/22 | `raw/2019-10-21_22/` | u-blox GNSS (vehicle signals in `processed/` and the table) | aggregated only | SPaT logs 13:00-19:00, segment speeds |\n\n## The ML-ready tables (`speed-prediction/`)\n\nOne row per sample along each trip, sampled by distance travelled; `trip` numbers the laps.\nColumns: the ego vehicle's position, speed and engine speed, plus the accelerator pedal in October\nand engine load and throttle in August (all readable over OBD-II), then `signal_k_*` for the next five signals along\nthe route (index, distance to it, current phase), `segment_*` (current segment, its measured\nspeed, distance remaining), `v2v_mean_speed` and `v2v_sigma_speed` (statistics of the volunteer\ntraces at this location) and, in August only, `lead_*` (lead vehicle offset and speed from the\naftermarket ADAS sensors). Missing values are NaN. Units follow the source: SI unless stated, segment speeds in mph, time in seconds from\nmidnight. Phase numbers follow Fort Collins' standard (1 NB left, 2 SB through, 3 WB left,\n4 EB through, 5 SB left, 6 NB through, 7 EB left, 8 WB through).\n\n```python\nfrom datasets import load_dataset\nds = load_dataset(\"AD-Assurance-Lab/connected-eco-driving-doe-ee0008468\", \"speed-prediction\")\n```\n\n## Everything else\n\n- `raw/`: the logs as recorded. ADAS logs are Vector `.blf` with MATLAB/CSV decodes; logger\n  CSVs from the volunteer cars use Freematics' format (time, PID, value; PID A latitude, B\n  longitude).\n- `processed/`: the team's synchronized and interpolated MATLAB structures per day, and\n  `processed/nn-training/` with the cleaned traces used to train the first networks.\n- `results/`: velocity-prediction results (October data).\n- `video/2019-05-15_zed-stereo.svo`: ZED st","encoding":"utf8","key":"README.md","name":"README.md","size":5250,"truncated":true}],"sourceObjectKey":"processed/nn-training/D01Cleaned.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/AD-Assurance-Lab/connected-eco-driving-doe-ee0008468/processed/nn-training/D01Cleaned.csv\")\ndf.head()","starterSamplePath":"maps/FtCoMap.PNG","totalObjects":129},"summary":"Real-world connected-vehicle data from urban drives in Fort Collins, Colorado, in 2019, collected for vehicle velocity prediction and predictive energy management under U.S.","title":"Connected Eco-Driving, Fort Collins 2019: Synchronized Vehicle, ADAS, V2V and Traffic-Signal (V2I) Data for Velocity Prediction (DOE EE0008468)","updatedAt":"2026-11-22T01:00:28.233Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/AD-Assurance-Lab/connected-eco-driving-doe-ee0008468"}}