An open-source remote air quality sensing device made by OPEnS Lab OSU. The device logs air quality parameters to the MongoDB database.
| Name | Role | |
|---|---|---|
| Quinn Yockey | Team lead, embedded software | yockeyq@oregonstate.edu |
| Shion Britten | Electrical & power analysis | brittesh@oregonstate.edu |
| Rowan Bignell | Software & Loom integration | bignellr@oregonstate.edu |
Wildfires across Oregon, Washington, and California pose a significant threat to the wine industry through smoke taint, which negatively impacts wine quality. Accurate, high-resolution data on smoke exposure at the vineyard level is critical for risk assessment, yet commercial devices can be high cost. To address this, OPEnS Lab has developed Wisp, an Arduino-based air quality monitoring device:
- Data is being used with analysis of actual grape samples in effort to model smoke taint risk
- Measures 12 air quality parameters including particulate matter, CO₂, CO, SO₂
- Solar battery charger and 4G telemetry allows use in remote locations
- Low-cost (under $1500) compared to commercial competitors
- Open-source software and hardware
Figure 1 (Left): Deployed Wisp v2 Unit
Figure 2 (Right): Deployed Wisp v1 Unit
Figure 3: Wisp block diagram
Due to the remote nature of most vineyard deployments, a convenient power source is usually not present. A solar panel is used to recharge batteries, allowing indefinite operation as long as there is adequate sunlight.
Each Wisp device is powered by a Feather M0+ microcontroller break-out board. Data is sampled from each sensor at a configurable interval, typically 5 minutes. The data is then saved to an SD card in CSV format for future analysis. Precise sleep control and SD data logging are made possible by the Hypnos v3.3 board.
If available, Wisp devices can be configured to upload measured data to a centralized MongoDB database over a 4G LTE or WiFi network. Data is sent using the MQTT protocol, allowing computer applications to subscribe to MQTT data stream feeds. To save power, batches of sample data are accumulated and transmitted at a set interval, typically every 6 hours. By aggregating data remotely, researchers can monitor environmental conditions and unit health across all deployment sites simultaneously. This permits convenient near-real-time analysis of measured data without physically gathering the data from the SD card.
Like the V1, the Wisp V2 measures PM1, PM2.5, PM4, PM10, temperature, humidity, VOC, and NOx.
The Wisp V2 improves upon the V1 by adding support for DFRobot's suite of MultiGas sensors for 12 different contaminants. For vineyard deployments, the V2 uses the O₃ (SEN0472), CO (SEN0466), and SO₂ (SEN0470) sensors from DFRobot. The particulate matter sensor was upgraded from the Sensirion SEN55 to the SEN66, providing readings with much lower noise as well as allowing for measurement of CO₂.
The V2 also features a complete case redesign, using a fully 3D-printed waterproof enclosure that improves access to the SD card. The battery capacity is expanded from three 10050 mAh LiPo batteries to five, significantly improving operational runtime.
Work on the V2 began in 2025 and was completed by the start of the 2026 fire season.
| Specification | Sensor | Repeatability | Accuracy | Full Range |
|---|---|---|---|---|
| Ambient Temperature | SHT31 | ±0.07 °C | ±0.3 °C | -40 to 125 °C |
| Humidity | SHT31 | ±0.015 %RH | ±2 %RH | 0 to 100 %RH |
| Carbon Dioxide (CO₂) | SEN66 | ±10 ppm | ±(50 ppm + 2.5% m.v.) | 0 to 40000 ppm |
| Specification | Sensor | Repeatability | Device Variation | Full Range |
| Volatile Organic Compounds (VOC) | SEN66 | ±5 VOC index or ±5% m.v. | ±15 VOC index or ±15% m.v. | 1 to 500 VOC index |
| Nitrogen Oxides Index (NOx) | SEN66 | ±10 NOx index or ±10% m.v. | ±50 NOx index or ±50% m.v. | 1 to 500 NOx index |
| Specification | Sensor | Max Drift | Precision | Full Range |
| Particulate Matter 1.0 (PM1) | SEN66 | ±1.25 μg/m³/yr to ±1.25 % m.v./yr | ±(5 μg/m³ + 5% m.v.) | 0 to 1000 μg/m³ |
| Particulate Matter 2.5 (PM2.5) | SEN66 | ±1.25 μg/m³/yr to ±1.25 % m.v./yr | ±(5 μg/m³ + 5% m.v.) | 0 to 1000 μg/m³ |
| Particulate Matter 4.0 (PM4) | SEN66 | ±1.25 μg/m³/yr to ±1.25 % m.v./yr | ±25 μg/m³ | 0 to 1000 μg/m³ |
| Particulate Matter 10 (PM10) | SEN66 | ±1.25 μg/m³/yr to ±1.25 % m.v./yr | ±25 μg/m³ | 0 to 1000 μg/m³ |
| Specification | Sensor | Resolution | Accuracy | Full Range |
| Carbon Monoxide (CO) | SEN0466 | 1 ppm | ±10% | 0 to 1000 ppm |
| Sulfur Dioxide (SO₂) | SEN0470 | 0.1 ppm | ±10% | 0 to 20 ppm |
| Ozone (O₃) | SEN0472 | 0.1 ppm | ±10% | 0 to 10 ppm |
Table 1: Wisp v2 Sensor specifications, '% m.v.' means '% of measured value'.
Like the Wisp V2 units, each Wisp V1 unit can measure particulate matter, volatile organic compounds, nitrogen oxides, temperature and humidity.
| Specification | Sensor | Resolution | Accuracy | Full Range |
|---|---|---|---|---|
| Ambient Temperature | SHT31 | 0.015 °C | ±0.3 °C | -40 to 125 °C |
| Humidity | SHT31 | 0.015 %RH | ±2 %RH | 0 to 100 %RH |
| Specification | Sensor | Repeatability | Device Variation | Full Range |
| Volatile Organic Compounds | SEN55 | ±5 VOC index or ±5% m.v. | ±15 VOC index or ±15% m.v. | 1 to 500 VOC index |
| Nitrogen Oxides Index | SEN55 | ±10 NOx index or ±10% m.v. | ±50 NOx index or ±50% m.v. | 1 to 500 NOx index |
| Specification | Sensor | Max Drift | Precision | Full Range |
| Particulate Matter 1.0 | SEN55 | ±1.25 μg/m³/yr to ±1.25 %m.v./yr | ±(5 μg/m³ + 5% m.v.) | 0 to 1000 μg/m³ |
| Particulate Matter 2.5 | SEN55 | ±1.25 μg/m³/yr to ±1.25 %m.v./yr | ±(5 μg/m³ + 5% m.v.) | 0 to 1000 μg/m³ |
| Particulate Matter 4 | SEN55 | ±1.25 μg/m³/yr to ±1.25 %m.v./yr | ±25 μg/m³ | 0 to 1000 μg/m³ |
| Particulate Matter 10 | SEN55 | ±1.25 μg/m³/yr to ±1.25 %m.v./yr | ±25 μg/m³ | 0 to 1000 μg/m³ |
Table 2: Wisp V1 Sensor specifications, '% m.v.' means '% of measured value'.
Figure 4: Fully built Wisp V1 device
The Pelican case has three holes drilled on the side to accommodate the PG7 cable glands and waterproof cable set. Inside the case, a custom 3D printed base plate holds electronics securely in place.
For the past four years, Wisp units have monitored air quality at 44 vineyards in Washington, Oregon, and California. This data is currently being analyzed by researchers at Oregon State University, Washington State University Tri-Cities, and University of California Davis.