Grow Room Sensor Monitoring Software: What to Track and Where to Place Sensors

Every commercial flower room already has sensors somewhere: a wall-mounted controller probe, a handheld meter clipped to a trellis, maybe a CO2 sensor tied to an enrichment system. The problem is usually not a shortage of sensors. It is that the readings do not automatically mean anything. A wall probe outside the canopy does not necessarily represent canopy air. A CO2 sensor next to a busy door can be biased by traffic and outdoor air.

Sensors provide inputs: a reading at a point in space and time. They do not know whether a CO2 dip came from an open door, a delivery fault, or the room’s operating plan, and they cannot explain why a reading moved unless someone connects it to the work that happened in the room that day. That connection between a graph and an operating record is what turns sensor data into a decision instead of a data point sitting in a dashboard.

This is not a hardware shopping guide. It walks through what a commercial cultivation environment needs to measure, where to put the hardware so readings mean something, how to spot drift before it distorts decisions, and how sensor data should connect to the rest of the operating record instead of living in an isolated app. None of this assumes new equipment. Most facilities already own working temperature, humidity, and light sensors, and the usual gap is placement, calibration discipline, and whether anyone ties a reading back to what the plant did afterward. Growgoyle can bring readings from existing sensor systems into the operating record through CSV import or an API, so getting started does not require replacing working hardware.

What a Commercial Room Actually Needs to Track

Five categories of data matter for day-to-day cultivation decisions: air temperature and relative humidity (which combine into vapor pressure deficit, or VPD), CO2 concentration, light intensity, root-zone water content and electrical conductivity (EC), and irrigation or runoff chemistry. Each answers a different operational question and has its own placement and verification requirements. Treating them as one generic “environment score” hides what is actually off when a batch underperforms.

Air temperature and RH are the most tracked and the most often mislocated. A crop-process CO2 sensor shows enrichment conditions at its location, but it is not automatically a worker-exposure monitor. Light data starts with photosynthetic photon flux density (PPFD); logged PPFD over time is used to calculate daily light integral (DLI). Root-zone data tells you what is happening in the substrate, which can be disconnected from what a wall-mounted air sensor reports. For the full picture of how these systems interact across a facility, see our cannabis climate control guide.

Air Conditions: Where Room Air and Leaf Conditions Diverge

Place the primary temperature and RH sensor in the plant zone at or near canopy height, with representative airflow and protection from direct solar or fixture radiation, irrigation water, and local heating or cooling sources. That is the greenhouse placement guidance summarized by Both et al. (2015). In a tall or fast-changing canopy, the sensing position may need to move as the crop moves. A representative air reading is still not the same as leaf temperature.

Leaf temperature can be above or below air temperature depending on radiation, transpiration, and air movement. The e-GRO VPD brief shows why the direction matters: in its worked example, a leaf 2°C cooler than room air had lower leaf VPD than air-only VPD. A fixed cooler-leaf assumption is not universal, so the operator should measure rather than automatically subtracting a degree or two. We cover this distinction in our leaf VPD versus air VPD guide and explain why CO2 and LED fixtures do not create a universal chart correction in our CO2 and LED VPD article. The broader plant-physiology literature also treats the leaf-to-air vapor pressure gradient as a driver of transpiration and stomatal response (Grossiord et al., 2020, New Phytologist).

The practical takeaway is not that every room needs a continuous leaf-temperature sensor. It is that an operator should spot-check representative leaves with an infrared thermometer when lighting, airflow, irrigation status, or canopy structure could change the leaf-to-air offset. Our VPD calculator lets you use a measured leaf temperature instead of a fixed offset.

CO2: Placement, Enrichment, and the Safety Line

A crop-process CO2 sensor should sample a representative point in or near the canopy, away from obvious local bias such as an injection outlet, doorway, or poorly mixed pocket. Both et al. (2015) also recommend enough sensing points to capture spatial variation rather than assuming one location represents a non-uniform room. Calibrate each sensor by the manufacturer’s procedure and service interval. Do not assume a fresh-air reset is valid for every NDIR sensor or for a room that rarely reaches outdoor background concentration.

CO2 enrichment is also a worker-safety question, not only a plant-growth question. OSHA’s permissible exposure limit is 5,000 ppm as an eight-hour time-weighted average (OSHA Chemical Sampling Information, Carbon Dioxide). Oklahoma State’s greenhouse guidance recommends a 2,000 ppm alarm for enrichment systems, but that crop-process recommendation is not a complete worker-safety design. Safety-rated monitors, alarm locations and setpoints, ventilation, and emergency response should follow the facility’s hazard assessment, equipment instructions, applicable code, and qualified safety guidance. A canopy-control probe should not be treated as the only proof that every work area is safe.

We go deeper on how CO2 enrichment interacts with VPD and lighting choices in our CO2 and LED VPD article, which is worth reading before setting enrichment targets for a room running supplemental lighting.

Light: PPFD, DLI, and Sensor Drift

Photosynthetic photon flux density and daily light integral are standard horticultural lighting quantities defined in ANSI/ASABE S640. A quantum sensor measures PPFD; a logger integrates those measurements over time to calculate DLI. A lux meter is weighted to human vision and is not a substitute for a quantum sensor when measuring light for plant growth. At canopy level, PPFD measurements capture fixture output, dirt, geometry, and shading that a fixture’s rated output does not.

Placement matters too. The sensor should sit level, clear of overhead obstructions that could shade it intermittently, and away from irrigation overspray, since residue on the sensor can drive an erroneously low reading (Runkle, “Proper Use and Maintenance of a Light Sensor”). Because canopy height and fixture geometry can change through a flowering cycle, map or spot-check PPFD at representative canopy points instead of treating a fixed wall or ceiling sensor as a canopy map.

Quantum sensor readings also deserve periodic service checks. Contamination, damage, aging, or a changed mounting position can bias the result, and a low reading does not identify its own cause. Follow the maker’s cleaning and calibration guidance, record the as-found condition, and compare instruments only when the reference itself has a known status.

Root Zone: Substrate Water Content, EC, and Runoff Checks

Air and canopy sensors tell you about the environment around the plant. Substrate sensors and extraction or runoff checks tell you about the root zone. Capacitance-based sensors infer volumetric water content from dielectric permittivity, and some models also report bulk EC. METER Group’s TEROS 12 documentation says its standard calibration is intended for mineral soils and that soilless substrates require a substrate-specific calibration. Do not assume one factory curve transfers cleanly among rockwool, coco coir, and peat-based media.

For container media, pour-through is one manual extraction method: irrigate the crop, wait for the documented equilibration period, add a specified volume of distilled water, collect enough leachate, and test pH and EC (UMass Extension). Use one published protocol consistently because pot size, solution volume, timing, and temperature affect the result. A pour-through value is not interchangeable with every runoff measurement or with a probe’s bulk-EC reading, but repeated measurements can provide a useful independent trend check.

Calibration and Comparison Checks You Can Run Without New Hardware

None of the sensor types above are maintenance-free, and most do not diagnose their own drift. Build each check around the manufacturer’s instructions, the facility’s risk assessment, and a documented reference:

• Compare controller air sensors with a traceable or recently verified reference instrument, and log the as-found difference before making an adjustment.
• Calibrate CO2 sensors by the maker’s procedure. Use the specified gas or baseline condition instead of assuming every NDIR sensor can be reset in fresh air.
• Map or spot-check PPFD at representative canopy points after fixture, canopy, or room-layout changes and at the service interval chosen for the facility.
• Clean and inspect light-sensor surfaces according to the maker’s guidance, especially where irrigation residue or dust can accumulate.
• Run a consistent root-zone extraction or runoff protocol as an independent trend check, while keeping its values distinct from probe readings that use a different measurement basis.

This belongs in the same operating record as everything else a crew does that day, not in a separate maintenance binder nobody opens between harvests.

From Readings to Decisions: Why Sensor Data Needs an Operating Record

A sensor dashboard, no matter how complete, shows you conditions. It does not show you what a crew did that day, what stage each room was in, or what actually happened at harvest three weeks later. Connecting a CO2 dip or a VPD swing to a batch outcome requires someone to log what happened in the room and review it against the readings later. Growgoyle’s searchable history brings notes and tasks back into that review, while its correlation tools can surface possible relationships among crop conditions, work, observations, sensor readings, and outcomes by flowering batch. A correlation points to something worth investigating. It does not establish cause. That gap between raw environmental data and an operating record with context is the difference between a sensor dashboard and something closer to cultivation intelligence, a distinction we cover directly in sensor dashboards versus cultivation intelligence.

Sensors do not replace an operator’s judgment about what a specific batch needs. They give that operator better information, provided the sensor is in the right place, checked on a known schedule, and its readings are reviewed alongside the work log and batch record rather than left to stand alone. Growgoyle can schedule and assign a calibration or comparison check, then retain the completed task and its note in searchable history. Environmental readings remain in the room and batch record. The interpretation stays with the person who knows the crop. Our cultivation task management guide covers how readings and observations become assigned work, while the grow room maintenance schedule covers recurring equipment checks.

Placement Checklist

• Place primary air temperature and RH sensors in the plant zone at or near canopy height, with representative airflow and protection from direct radiation, irrigation, and local heating or cooling sources.
• Use enough sensing points to characterize large or non-uniform rooms instead of assuming one point represents the whole space.
• Put the crop-process CO2 sensor at a representative canopy location, away from injection outlets, busy doors, and poorly mixed pockets.
• Treat worker-exposure monitoring as a separate safety design based on a hazard assessment, equipment instructions, and applicable requirements.
• Map or spot-check PPFD at representative canopy points and after material changes in canopy or fixture geometry.
• Keep light-sensor surfaces clean and inspect them by the maker’s procedure.
• Use a substrate-specific calibration when the probe’s standard curve does not cover the production medium.
• Use one documented root-zone extraction or runoff protocol consistently and do not equate unlike EC measurement methods.
• Log each calibration and comparison check against the room and date, including the as-found difference and action taken.
• Tie environmental readings to the batch and work record so a reading has context when a batch is reviewed later.

References

1. Both, A. J., Benjamin, L., Franklin, J., et al. (2015). “Guidelines for Measuring and Reporting Environmental Parameters for Experiments in Greenhouses.” Plant Methods, 11, 43. https://pmc.ncbi.nlm.nih.gov/articles/PMC4567830/
2. Runkle, E. “Proper Use and Maintenance of a Light Sensor.” Greenhouse Product News, Michigan State University Department of Horticulture. https://gpnmag.com/article/greenhouse-light-sensor-best-practices/
3. Oklahoma State University Extension. “Greenhouse Carbon Dioxide Supplementation.” HLA-6723, 2017. https://extension.okstate.edu/fact-sheets/print-publications/hla/greenhouse-carbon-dioxide-supplementation-hla-6723.pdf
4. Occupational Safety and Health Administration. “Carbon Dioxide” Chemical Sampling Information. https://www.osha.gov/chemicaldata/183
5. UMass Extension. “Soil Testing,” section on the pour-through method for container media. https://www.umass.edu/agriculture-food-environment/greenhouse-floriculture/fact-sheets/soil-testing
6. METER Group. TEROS 12 product documentation and calibration guidance. https://metergroup.com/products/teros-12/
7. e-GRO Alert. “VPDleaf vs. VPDair: Two Different Ways to Determine VPD.” https://e-gro.org/pdf/e816.pdf
8. Grossiord, C., Buckley, T. N., Cernusak, L. A., et al. (2020). “Plant Responses to Rising Vapor Pressure Deficit.” New Phytologist. https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.16485
9. American Society of Agricultural and Biological Engineers. ANSI/ASABE S640, Quantities and Units of Electromagnetic Radiation for Plants (Photosynthetic Organisms). https://elibrary.asabe.org/azdez.asp?JID=2&AID=48303&CID=s2000&T=2