C6H6(GT)fev / AutoGluon ↗UCI Air Quality Chemical Gas Sensor Telemetry
Multivariate environmental time series benchmark capturing 9,357 hourly observations from a chemical multisensor array deployed in an Italian urban road network. The task evaluates multi-target hourly forecasting of critical air pollutants (Benzene, CO, NO2, NOx) across a 168-hour (7-day) operational planning horizon.
Business Objective: UCI Air Quality Chemical Gas Sensor Telemetry
01Business Context
Municipal environmental protection agencies (EPAs) and smart city traffic dispatchers require rolling 7-day forecasts of toxic urban pollutants to schedule industrial traffic restrictions, alert vulnerable populations, and manage urban ventilation corridors.
02Analytical Objective
Predict hourly atmospheric concentrations of ground-truth Benzene (C6H6), Carbon Monoxide (CO), Nitrogen Dioxide (NO2), and Total Nitrogen Oxides (NOx) over a 168-hour horizon H using multisensor array resistance and weather covariates.
03Economic & Decision Impact
Under-forecasting risks severe public health exposure events, WHO particulate exceedances, and delayed municipal smog emergencies; over-forecasting triggers unnecessary low-emission zone traffic shutdowns and costly commercial transport diversions.
Classical univariate baselines (AutoARIMA, Seasonal Naive) struggle with the high cross-series covariance and non-linear chemical sensor cross-sensitivities under shifting humidity and temperature. Gradient boosted lag ensembles (LightGBM) provide strong predictive baselines when supplied with rolling weather windows, while modern zero-shot foundation models (Mistral's Timestral, MOIRAI, TimesFM) excel at zero-shot transfer across the 168-step horizon by leveraging multi-target temporal cross-attention.
Target & Feature Column Definitions
Exhaustive business definitions, measurement units, roles, and target variables across all 15 columns.
| Col | Variable / Feature Name | Role | Type | Unit / Scale | Description & Business Meaning |
|---|---|---|---|---|---|
| C | C6H6(GT)Benzene C6H6 (Ground Truth) | TARGET | NUM | µg/m³ | Reference analytical concentration of Benzene measured by standard reference analyzer |
| D | CO(GT)Carbon Monoxide CO (Ground Truth) | TARGET | NUM | mg/m³ | Reference analytical concentration of Carbon Monoxide measured by standard reference analyzer |
| E | NO2(GT)Nitrogen Dioxide NO2 (Ground Truth) | TARGET | NUM | µg/m³ | Reference analytical concentration of Nitrogen Dioxide measured by standard reference analyzer |
| F | NOx(GT)Nitrogen Oxides NOx (Ground Truth) | TARGET | NUM | ppb | Reference analytical concentration of Nitrogen Oxides measured by standard reference analyzer |
| A | idMonitoring Station ID | SERIES ID | CAT | — | Unique categorical identifier for the urban sensor installation site |
| B | timestampObservation Timestamp | TIME | TIME | — | ISO 8601 hourly date and time of air sample collection |
| G | AHAbsolute Humidity | Feature | NUM | g/m³ | Atmospheric absolute humidity influencing chemical sensor surface conductance |
| H | RHRelative Humidity | Feature | NUM | % | Atmospheric relative humidity percentage |
| I | TAmbient Temperature | Feature | NUM | °C | Ambient air temperature at measurement site |
| J | PT08.S1(CO)PT08.S1 Tin Oxide CO Sensor | Feature | NUM | Ω | Hourly nominal resistance sensor response nominally targeted to Carbon Monoxide |
| K | NMHC(GT)Non-Methane Hydrocarbons (Ground Truth) | Feature | NUM | µg/m³ | Reference analytical concentration of Non-Metanic HydroCarbons |
| L | PT08.S2(NMHC)PT08.S2 Titania NMHC Sensor | Feature | NUM | Ω | Hourly nominal resistance sensor response nominally targeted to Non-Methane Hydrocarbons |
| M | PT08.S3(NOx)PT08.S3 Tungsten Oxide NOx Sensor | Feature | NUM | Ω | Hourly nominal resistance sensor response nominally targeted to Nitrogen Oxides |
| N | PT08.S4(NO2)PT08.S4 Tungsten Oxide NO2 Sensor | Feature | NUM | Ω | Hourly nominal resistance sensor response nominally targeted to Nitrogen Dioxide |
| O | PT08.S5(O3)PT08.S5 Indium Oxide Ozone Sensor | Feature | NUM | Ω | Hourly nominal resistance sensor response nominally targeted to Ozone |
Interactive Data Table
Explore rows, feature values, and target labels for UCI Air Quality Chemical Gas Sensor Telemetry.
Foundation Models vs Traditional ML & Baselines (41 Methods)
Empirical evaluation directly from the AutoGluon fev benchmark suite evaluating zero-shot foundation models against deep learning, gradient-boosted trees, and statistical baselines across 100 evaluation tasks.
| Rank | Method / Model | Family | Organization | SQL Win Rate?Average win rate (%) against all competing models across the benchmark tasks under Scaled Quantile Loss. | SQL Skill?Normalized skill score relative to the Seasonal Naive baseline (0% = baseline, higher is better). | MASE Win Rate?Average win rate (%) under Mean Absolute Scaled Error. | Latency?Median end-to-end execution time per 100 time series in seconds. | Link |
|---|---|---|---|---|---|---|---|---|
| #1 | Timestral-209M | Foundation Model | Mistral AI | 87.7% | +50.4% | 86.0% | 0.51s | Repo ↗ |
| #2 | Timestral-123M (Carla Engine) | Foundation Model | Mistral AI | 86.8% | +50.2% | 85.0% | 0.32s | Repo ↗ |
| #3 | TimeRSI | Closed-Source API | GaTech & SamayLabs | 84.8% | +51.2% | 83.8% | 2512s | Repo ↗ |
| #4 | t0-gamma | Foundation Model | The Forecasting Company | 84.4% | +49.4% | 81.8% | 4.38s | Repo ↗ |
| #5 | Timestral-28M | Foundation Model | Mistral AI | 79.2% | +48.4% | 77.0% | 0.11s | Repo ↗ |
| #6 | timesfm-3 | Foundation Model | 78.5% | +49.0% | 77.0% | 3.41s | Repo ↗ | |
| #7 | ChakraTS | Baseline | Community | 78.2% | +48.1% | 79.5% | 14.22s | Repo ↗ |
| #8 | Lingjiang2.0 | Baseline | Community | 73.9% | +47.0% | 71.3% | 8.89s | Repo ↗ |
| #9 | Chronos-2 | Foundation Model | AWS | 70.4% | +47.3% | 68.4% | 0.84s | Repo ↗ |
| #10 | timeh2o-v1-tiny | Foundation Model | H2O.ai | 69.9% | +45.5% | 67.0% | 0.19s | Repo ↗ |
| #11 | Timestral-5M | Foundation Model | Mistral AI | 69.0% | +46.2% | 67.7% | 40ms | Repo ↗ |
| #12 | Toto-2.0-2.5B | Baseline | Community | 67.5% | +44.4% | 65.7% | 5.58s | Repo ↗ |
| #13 | t0-beta | Foundation Model | The Forecasting Company | 67.4% | +46.6% | 64.5% | 2.29s | Repo ↗ |
| #14 | Toto-2.0-1B | Baseline | Community | 66.4% | +44.4% | 65.5% | 2.48s | Repo ↗ |
| #15 | TiRex-2 | Foundation Model | NX-AI | 65.8% | +45.5% | 63.4% | 0.27s | Repo ↗ |
| #16 | Toto-2.0-313m | Baseline | Community | 65.1% | +44.2% | 63.5% | 1.06s | Repo ↗ |
| #17 | TiRex | Foundation Model | NX-AI | 58.0% | +42.6% | 54.1% | 0.24s | Repo ↗ |
| #18 | Toto-2.0-22m | Baseline | Community | 57.6% | +42.9% | 56.0% | 0.25s | Repo ↗ |
| #19 | TimesFM-2.5 | Baseline | Community | 54.6% | +42.2% | 53.5% | 1.87s | Repo ↗ |
| #20 | TabPFN-TS-3 | Foundation Model | Prior Labs | 52.4% | +43.1% | 51.3% | 235s | Repo ↗ |
| #21 | TS-ICL | Foundation Model | EDF-Lab | 51.2% | +43.1% | 49.8% | 2.49s | Repo ↗ |
| #22 | FlowState | Foundation Model | IBM | 49.3% | +39.9% | 49.9% | 2.29s | Repo ↗ |
| #23 | citras-fm | Foundation Model | Hitachi | 48.8% | +41.2% | 46.9% | 0.60s | Repo ↗ |
| #24 | Toto-1.0 | Baseline | Community | 47.5% | +40.8% | 45.9% | 22.35s | Repo ↗ |
| #25 | TabPFN-TS | Foundation Model | Prior Labs | 47.3% | +41.5% | 45.4% | 109s | Repo ↗ |
| #26 | Toto-2.0-4m | Baseline | Community | 46.2% | +40.7% | 43.6% | 0.20s | Repo ↗ |
| #27 | Moirai-2.0 | Baseline | Community | 41.9% | +39.3% | 40.5% | 0.35s | Repo ↗ |
| #28 | Chronos-Bolt | Foundation Model | AWS | 41.4% | +38.9% | 39.3% | 0.26s | Repo ↗ |
| #29 | TFT | Deep Learning | — | 31.7% | +32.2% | 29.2% | 829s | Repo ↗ |
| #30 | Stat. Ensemble | Baseline | Community | 27.5% | +20.7% | 30.6% | 147s | Repo ↗ |
| #31 | Sundial-Base | Foundation Model | Tsinghua University | 27.4% | +33.4% | 33.6% | 8.01s | Repo ↗ |
| #32 | PatchTST | Deep Learning | — | 26.8% | +30.1% | 25.1% | 664s | Repo ↗ |
| #33 | DeepAR | Deep Learning | — | 26.1% | +29.1% | 23.6% | 1113s | Repo ↗ |
| #34 | AutoARIMA | Statistical Baseline | — | 23.3% | +21.3% | 21.5% | 20.14s | Repo ↗ |
| #35 | AutoETS | Statistical Baseline | — | 20.3% | -26.8% | 22.5% | 3.48s | Repo ↗ |
| #36 | CatBoost | Tree Ensemble | — | 19.1% | +23.0% | 35.1% | 29.48s | Repo ↗ |
| #37 | LightGBM | Tree Ensemble | — | 17.8% | +21.0% | 33.1% | 3.03s | Repo ↗ |
| #38 | AutoTheta | Statistical Baseline | — | 15.6% | +5.5% | 19.8% | 3.28s | Repo ↗ |
| #39 | Seasonal Naive | Baseline | Community | 10.3% | 0.0% | 11.8% | 0.48s | Repo ↗ |
| #40 | Naive | Statistical Baseline | — | 6.7% | -45.4% | 10.9% | 0.47s | Repo ↗ |
| #41 | Drift | Statistical Baseline | — | 6.1% | -45.8% | 10.0% | 0.45s | Repo ↗ |
Variable Schema & Summary Distributions
Observed numerical ranges, category cardinalities, missing rates, and sample values across 9,357 rows.
| Col | Variable Name | Type | Missing | Distinct | Summary Stats / Distribution | Sample Values |
|---|---|---|---|---|---|---|
| C | C6H6(GT)TARGET | NUM | 366 (0.03912%) | 407 | [0.10000000149011612, 63.70000076293945] μ=10.1 σ=7.4 | 11.8999996185302739.3999996185302739 |
| D | CO(GT)TARGET | NUM | 1,683 (0.17987%) | 96 | [0.10000000149011612, 11.899999618530273] μ=2.2 σ=1.5 | 2.599999904632568422.200000047683716 |
| E | NO2(GT)TARGET | NUM | 1,642 (0.17548%) | 283 | [2, 340] μ=113.1 σ=48.4 | 11392114 |
| F | NOx(GT)TARGET | NUM | 1,639 (0.17516%) | 925 | [2, 1479] μ=246.9 σ=213.0 | 166103131 |
| A | id | CAT | 0 | 1 | — | 000 |
| B | timestamp | CAT | 0 | 9,357 | — | 2004-03-10T18:00:002004-03-10T19:00:002004-03-10T20:00:00 |
| G | AH | NUM | 366 (0.03912%) | 6,683 | [0.18469999730587006, 2.2309999465942383] μ=1.0 σ=0.4 | 0.75779998302459720.72549998760223390.7501999735832214 |
| H | RH | NUM | 366 (0.03912%) | 753 | [9.199999809265137, 88.69999694824219] μ=49.2 σ=17.3 | 48.90000152587890647.7000007629394554 |
| I | T | NUM | 366 (0.03912%) | 436 | [-1.899999976158142, 44.599998474121094] μ=18.3 σ=8.8 | 13.60000038146972713.30000019073486311.899999618530273 |
| J | PT08.S1(CO) | NUM | 366 (0.03912%) | 1,041 | [647, 2040] μ=1099.8 σ=217.1 | 136012921402 |
| K | NMHC(GT) | NUM | 8,443 (0.90232%) | 429 | [7, 1189] μ=218.8 σ=204.5 | 15011288 |
| L | PT08.S2(NMHC) | NUM | 366 (0.03912%) | 1,245 | [383, 2214] μ=939.2 σ=266.8 | 1046955939 |
| M | PT08.S3(NOx) | NUM | 366 (0.03912%) | 1,221 | [322, 2683] μ=835.5 σ=256.8 | 105611741140 |
| N | PT08.S4(NO2) | NUM | 366 (0.03912%) | 1,603 | [551, 2775] μ=1456.3 σ=346.2 | 169215591555 |
| O | PT08.S5(O3) | NUM | 366 (0.03912%) | 1,743 | [221, 2523] μ=1022.9 σ=398.5 | 12689721074 |
Frequently Asked Questions: UCI Air Quality Chemical Gas Sensor Telemetry
Common questions regarding the UCI Air Quality Chemical Gas Sensor Telemetry dataset, machine learning task formulations, and in-browser tabular inference.
What is the forecasting horizon for UCI Air Quality Chemical Gas Sensor Telemetry?
The standard forecasting horizon is 168 hours (7 full days), matching weekly municipal environmental compliance planning and air quality alert windows.
Can I run zero-server forecasts on this dataset in Google Sheets?
Yes, Carla enables zero-server in-context time series forecasting directly in spreadsheets using Mistral's Timestral foundation model without moving data to an external server.
Which models perform best on UCI Air Quality Chemical Gas Sensor Telemetry?
Zero-shot foundation models such as Mistral's Timestral and lag-augmented LightGBM models deliver top performance by modeling non-linear temperature and humidity interactions with sensor telemetry.
Test TabICLv2 on UCI Air Quality Chemical Gas Sensor Telemetry Yourself
Open the pre-loaded Google Sheet and let Carla configure the target and task for local, zero-cloud tabular machine learning.
Launch Carla
Click once to open the spreadsheet and Carla side panel together.
Review the Setup
Carla selects the dataset target and task from this page automatically.
Evaluate & Predict
Run predictions and compute metrics with zero server uploads.
Provenance & Attribution
De Vito, S., Massera, E., Piga, M., Martinotto, L., & Di Francia, G. (2008). On field calibration of an electronic nose for benzene estimation in an urban environment. Sensors and Actuators B: Chemical, 129(2), 750-757.