The VRWE automatic irrigation controller
HomePrecisionThe VRWE automatic irrigation controller
Precision · Automatic irrigation

The VRWE automatic irrigation controller

This paper shows how VRWE uses some signals together to calculate the water in the root zone more accurately than one probe. You will know the cause when VRWE waits and does not apply water. You will also know how to read the confidence of VRWE and if you can accept the estimate.

Precision9 diagrams5 sources~9 min to read
In this paper
01 Purpose and scope02 Definitions03 Evidence and limitations04 Some signals together correct one incorrect reading05 Confidence: how sure the estimate is06 Irrigation decisions07 How to read shared-drain measurements08 Troubleshooting09 Expected results and limitations10 References
Start here

Purpose and scope

Weak

Limit of the data: The method of VRWE is good. It uses more than one signal and it is careful. You cannot be sure that VRWE does not apply too much water or not sufficient water. The sensors must be in good condition, the calibration must be correct, and the fail-safe functions must operate. Keep the emergency minimum VWC values. Make sure that a person can override VRWE.

Virtual Root-Zone Water Estimator (VRWE) is a software controller that selects when to apply water to a plant and how much water to apply. VRWE uses some signals together and does not obey one sensor. The name includes ‘virtual’ because VRWE does not measure the water directly. VRWE calculates the quantity of water from some signals.

Each pot has only one moisture sensor. The sensor measures only a small part of the substrate[1]. If this part is dry, or if water flows around it, the sensor shows that the pot is dry. A controller that obeys only the sensor then applies too much water and can cause damage to the plant. VRWE uses the sensor reading as one signal that it examines. VRWE does not obey the sensor.

How VRWE operates: three steps1SENSORone sensor,small area, canbe incorrect→2ESTIMATEVRWE calculatesthe waterbalance→3DECISIONapply water /wait / tell apersonThe sensor is the first step. It is not the decision.
Figure 1. VRWE gets one signal from the sensor. The sensor measures a small area and can be incorrect. VRWE compares this signal with other data and calculates an estimate. Then VRWE makes a careful decision. All of this paper is about the ESTIMATE step.
About this paper

This paper gives information on how the system operates. This paper is not a report of a laboratory test. Read this paper with the root-zone sensor paper (the quantities that one probe measures) and the signal and noise paper (how to find if a change is in the plant or only sensor noise).

Terms

Definitions

Five terms are important in all parts of this paper. This section gives the five definitions first. It is not necessary to know them at this time, because each term occurs again in the sections that follow.

VWC (volumetric water content)The quantity of water in the substrate at the position of the sensor, as a percentage. VRWE compares this sensor reading with other data.
Runoff / drainWater that flows out of the bottom of the pot. The plant does not use this water.
Full pot (DUL, drained upper limit)The maximum quantity of water that the pot can hold after the drainage stops. The water that you add after this point flows out as runoff.
ChannelingWater that flows straight down one channel and does not touch the roots. An example is a gap between the pot and the liner. All the water uses this easy channel and not the substrate. The water flows out at the bottom, and the root zone near the channel stays dry. The water goes in at the top and out at the bottom, and it does not help the plant.
ConfidenceThe value that VRWE gives to show how accurate its estimate is at this time. High confidence lets VRWE apply more water. Low confidence makes VRWE careful.
The five terms that are important in this paper.
TermDefinition
VWCThe quantity of water in the substrate at the position of the sensor
Runoff / drainWater that flows out of the bottom of the pot
Full pot (DUL)The maximum quantity of water that the pot holds after the drainage stops
ChannelingWater that flows to the drain and does not touch the roots
ConfidenceThe value that shows how accurate the estimate of VRWE is at this time
Water-content cycle for one dayFill to field capacity, hold it, make a controlled dryback, and do it again. The dryback size is the steering control.correct bandtoo dry0255075100offP3 minonP1FCP2offwater content %
Diagram. The controller keeps the water-content curve for each day at the setpoints. It does not use a timer.
Saturation, field capacity and drybackthe same block at three times of the daySaturatedafter irrigation, almostno air→Field capacityfree water drained, peakeach day→Dryback lowplant used water, air +oxygen in
Diagram. A full pot (the maximum after drainage) and the dryback low give the band that the controller controls.
A block that is too dry cannot become wet again from the dripperMoist: it becomes wet again equallythe water goes to all parts of the fiberToo dry: water flows through channelsdry centerflows through channels, the center stays dryWhen the WC is less than ~30%, the dry fiber cannot pull water through the block. Water finds channels and flows out as runoff. The dripper cannot saturate a dry block again. Soak the block manually.
Diagram. Channeling: when a pot is too dry, the water flows in the same areas each time. The remaining substrate stays dry.
Quality of the information

Evidence and limitations

Three types of information

We want each paper to be correct. Thus we examine each paper for information that is not solid. This information comes from the view of one grower, from a small number of sources, or from grower methods that have no controlled test. We identify this information, and we do not show it as solid.

Frequently there is no paper for the decision that you must make. Then we use the reports of other growers and the methods that work in the rooms that we know. This information can help you, but it is not the result of a laboratory test. Use the methods that work for your plants, for your room and for your meters. If a table does not agree with your crop, use the data from your crop and record the difference.

Solid
Tests of plants, specifications or many sources that agree show that the information is correct
  • One moisture probe can give an incorrect reading. A check of more than one signal is good engineering.
Grower method
Many growers use this method in their rooms. Start with it, then change it for your room.
  • Sensor fusion of the VRWE type, as a safety structure for automatic irrigation.
Weak
The view of one grower, a small number of sources, one test, or a method that works in one room
  • A system that, in all failure modes, always applies a sufficient quantity of water and does not apply too much water.
  • Transpiration proxies as accurate models of the quantity of water that the crop uses, with no crop coefficients.

If you find an error, make a report. We will correct it. Make the report on GitHub, with the name of the paper and the information that is incorrect. If you have a source, include it: Make a report of an accuracy problem. The regulations, labels and licenses of your area always override each procedure in these papers. The label weak in a paper shows the points where the risk of an error is highest.

Primary information 1

Some signals together correct one incorrect reading

VRWE keeps a water balance and does not obey one probe. IN is the water that the drippers supply. The quantity is accurate because you calibrate the drippers. Thus you know accurately how much water you apply. OUT is the water that the plant uses and the runoff. The balance is the quantity of water that is in the pot.

A plant continuously absorbs water through its roots and releases the water as vapor through small pores in the leaves. When the temperature and the light intensity increase, the plant releases the vapor at a higher rate. Transpiration is the movement of water from the roots through the stems and out through the leaf pores as vapor.

You can measure temperature and light. Thus VRWE can calculate the rate at which the plant uses water, from these readings only[3]. As a result, VRWE has an estimate of OUT that does not use the sensor. VRWE compares the sensor reading with the balance. The sensor is not the only source of correct data.

One irrigation cycle as a water balanceEach cycle: add the water you apply. Subtract the plant uptake and the runoff.02142648555%Start75%+ IN (irrigation)66%- plant uptake60%- runoff
Figure 2. VRWE calculates the water balance again in each cycle. The last number is the best estimate of VRWE for the water in the root zone. VRWE gets this number before it examines the sensor reading.
accurate
Inputs IN
The volume of irrigation water that the calibrated drippers supply.
approximate
Inputs OUT
The plant uptake and the drainage. VRWE calculates the plant uptake from the temperature and the light.
check
One signal
The VWC reading of the sensor. VRWE compares the reading with other data and does not obey the reading.
output
Estimate + confidence
VRWE puts all the signals together in one number. It also gives the confidence value for the number.
Figure 3. Some signals go into one estimate. VRWE has more than one signal from different sources. Thus the other signals can correct one incorrect signal, and VRWE does not obey the incorrect signal.
Use more than one signal

If you use only one sensor, one incorrect reading becomes an incorrect decision. If the estimate uses some signals from different sources, the other signals correct one incorrect signal. The system stays correct when one input is incorrect.

Primary information 2

Confidence: how sure the estimate is

VRWE gives a confidence meter with each estimate. The meter shows how sure VRWE is. If you have only a number, for example ‘58% wet’, you do not know how accurate the number is.

The confidence is high when the signals from different sources agree. This occurs when the sensor, the water balance and the uptake estimate show the same result. The confidence decreases when the signals do not agree. For example, the sensor shows that the substrate is dry, but the water balance shows that the pot is full.

A defective sensor can only make VRWE more careful. The sensor cannot cause VRWE to find more room for water than the water balance shows. The water balance gives protection to the plant.

The confidence meter and the decision in each zoneHigher confidence lets VRWE apply more water. Low confidence makes VRWE careful.Low: wait / tell a personModerate: small safe shotHigh: apply a full shot0%50%100%
Figure 4. Confidence is a value on a scale. It is not a yes or no. A shot is the quantity of water that you apply in one irrigation. When the confidence is high, VRWE applies a full shot. When the confidence is moderate, VRWE applies only a small safe shot. When the confidence is low, VRWE waits or tells a person.
The system stays safe with a defective sensor
  • Each estimate has a confidence value, and not only a number.
  • When signals from different sources agree, the confidence increases. When they do not agree, the confidence decreases.
  • Low confidence makes VRWE careful. VRWE does not apply a full shot.
  • A defective sensor causes VRWE to wait. It does not cause too much water or not sufficient water.
Primary information 3

Irrigation decisions

VRWE makes one of only three decisions. The decision uses two values: the confidence and the headroom (the quantity of water that you can add before the pot is full).

The three decisions of VRWE1High confidence + headroomapply a fullshot→2Not surewait, or apply asmall safe shot→3Signals do not agreetell a personand make nodecisionIn all decisions, a small temporary deficit is better than too much water if VRWE is not sure. The emergency VWC minimum is also necessary.
Figure 5. VRWE makes only three decisions. In all three decisions, a small temporary deficit is better than too much water when VRWE is not sure.

When the confidence is high and there is headroom, VRWE applies a full shot. When VRWE is not sure, it waits or applies a small safe shot, and it does not apply a full shot. When the signals do not agree and VRWE cannot find which signal is correct, it tells a person. In all three decisions, a small temporary deficit of water is better than too much water if VRWE is not sure. The emergency minimum VWC is also necessary.

The safe decision

Apply more water only when the confidence is high. If VRWE is not sure, it makes the safe decision. This one instruction makes automatic irrigation safe.

In a facility

How to read shared-drain measurements

In a facility, the conditions are not as easy as for one pot. Frequently, three grow rooms drain into one shared sump (a bucket that collects the water, with a pump). The condensate from the air conditioner (AC) and from the dehumidifier also goes into the same bucket. The pump removes the water from the bucket in a pump cycle. After a pump cycle, you cannot know if the water is from too much irrigation in a room or only from the AC condensate. VRWE finds the cause in three steps.

  1. 1
    Find the background drip
    At night, irrigation is off. The only water that goes into the sump is the condensate from the AC and from the dehumidifier. VRWE finds the quantity of this stable background drip. Then VRWE subtracts it from each reading that follows.
  2. 2
    Use different irrigation times
    Each room has a different irrigation time. Thus the time of each pump cycle agrees with the irrigation time of one room. As a result, VRWE can find which room caused each pump cycle.
  3. 3
    Show ‘not sure’ when VRWE cannot find the cause
    If two irrigations occur at the same time, VRWE cannot find the cause of a pump cycle. Then VRWE shows ‘not sure’ for the reading and gives no cause. VRWE uses the same fail-safe method in all other parts of the system.
source
Room A
The room drains into the shared sump.
source
Room B
The room drains into the shared sump.
source
Room C
The room drains into the shared sump.
noise
AC + dehumidifier
A continuous drip of condensate. VRWE finds the drip at night and subtracts it.
problem
Shared sump + pump
One bucket and one pump cycle. You cannot know which room caused the cycle.
solution
Find the cause
Subtract the background drip. Use different irrigation times. Show ‘not sure’ when VRWE cannot find the cause.
Figure 6. Three rooms, the AC and the dehumidifier all send water to one sump. Thus you cannot know the cause of one pump cycle. The three steps find which room caused each pump cycle.
Different times show the cause of each pump cycleThe low baseline is the AC drip. Each peak occurs when one room receives irrigation.01225385000:00A irrigates+30mB startB irrigates+30mC startC irrigates+30mdrain rate
Figure 7. The rooms have different irrigation times. Thus each peak in the drain rate occurs at the time of the room that caused it. A pump cycle immediately after the irrigation of a room shows that the pots of the room are full or have too much water.
The information from a pump cycle after irrigation

A pump cycle immediately after the irrigation of a room shows that the pots of the room are at ‘full pot’ and that water flows out. You can use this information. It is not a fault.

When there is a problem

Troubleshooting

The problems below can occur. VRWE must continue to operate correctly when they occur. VRWE is the solution to these problems. The protection is the same for each problem. VRWE compares the reading with the water balance and decreases the confidence. VRWE does not do a task because of one reading that is possibly incorrect.

incorrect
Sensor in a dry area
The probe is in a local dry area and shows that the pot is dry. The remaining part of the pot has the correct water content.
incorrect
Channeling
Water flows quickly down one side of the pot. It goes around the sensor and the roots, straight to the drain.
incorrect
Shared-drain problem
One bucket does not show which room had too much water.
Figure 8. Three usual problems. In each problem, one signal is incorrect. None of the problems causes VRWE to make an incorrect decision. The water balance and the confidence meter show that the signals do not agree.

Look carefully at channeling, because it is not easy to find. In a container, water can flow in a preferential flow path, a fast channel that sends the irrigation water around the root zone[4]. The volume of irrigation water that you apply is correct, but the water does not touch the roots and shows only as drain. The properties of a soilless substrate change the speed at which water flows through the substrate and out of it[5]. Thus VRWE monitors the time of the drain and not only the volume of the drain.

in
Top of the pot
The irrigation water goes into the pot here.
around
Preferential flow path
Water flows quickly down one side. It goes around the sensor and the roots.
out
Drain
The water flows out at the bottom almost immediately. The roots stay dry.
Figure 9. A cross-section of channeling: the water goes in at the top and out at the bottom, and it does not touch the roots. In the water balance, a channel shows as a large quantity of water in and a large quantity of drain immediately after it. VRWE uses this sign to decrease the confidence.
The protection is always the same

VRWE does not obey a reading that is possibly incorrect. VRWE compares the reading with the water balance. If the two do not agree, VRWE decreases the confidence and is careful. One incorrect sensor reading does not cause too much water or not sufficient water for the plant.

The limits

Expected results and limitations

The results that VRWE gives and the results that it does not give
  1. The safe decision. VRWE applies more water only when the confidence is high. When the confidence is not high, VRWE makes the safe decision. VRWE calculates estimates from signals, and it cannot know more than the signals show.
  2. The worst result is that VRWE is too careful. A defective sensor makes VRWE careful. The defective sensor does not cause very bad damage. VRWE will wait or tell a person before it applies too much water or not sufficient water.
  3. It is usual that VRWE waits or tells a person. VRWE must make these decisions when it is not sure. They show that the system operates correctly. They do not show a fault.
  4. Incorrect input gives incorrect output. The estimate is only as good as the inputs. It is important that the volumes of the drippers are accurate and that VRWE has the drain baseline. Drift in the calibration of the sensor slowly decreases the accuracy of each estimate that uses the sensor.[2]

VRWE is slower than a timer by a small quantity, but it is much safer. At some times, VRWE waits and a controller with only a timer applies water. This difference is the correct operation of VRWE.

For more information about the measurements of one probe, read the root-zone sensor paper. VRWE finds if a change is in the plant or only sensor noise before it makes a decision. For information on this check, read the signal and noise paper.

Related papers
Root-zone condition · TEROS-12
How to read the values from the sensor
Signal and noise
Find the difference between plant changes and sensor noise
The closed loop
Levers, signals and plant state

References

  1. Szerement, J., Woszczyk, A., Szypłowska, A., Kafarski, M., Lewandowski, A., Wilczek, A., & Skierucha, W. (2019). A Seven-Rod Dielectric Sensor for Determination of Soil Moisture in Well-Defined Sample Volumes. Sensors, 19(7), 1646. https://doi.org/10.3390/s19071646
  2. Mane, S., Das, N., Singh, G., Cosh, M., & Dong, Y. (2024). Advancements in dielectric soil moisture sensor calibration: A comprehensive review of methods and techniques. Computers and Electronics in Agriculture, 218, 108686. https://doi.org/10.1016/j.compag.2024.108686
  3. Koehler, T., Wankmüller, F. J. P., Sadok, W., & Carminati, A. (2023). Transpiration response to soil drying versus increasing vapor pressure deficit in crops: physical and physiological mechanisms and key plant traits. Journal of Experimental Botany, 74(16), 4789-4807. https://doi.org/10.1093/jxb/erad221
  4. Owen, J., & Norden, D. (Profile Products). Understanding drainage in horticultural growing media. Greenhouse Management. (source not from a journal, for example from a manufacturer) https://www.greenhousemag.com/article/growing-media-defining-drainage-improve-substrate/
  5. International Society for Horticultural Science (ISHS). Utilizing the HYDRUS model as a tool for understanding soilless substrate water dynamics. Acta Horticulturae 1168. https://www.ishs.org/ishs-article/1168_41

Each citation in the paper, for example [1], refers to an item in this list. Each item is a primary source or an instruction from an authority, unless the item shows a different type. The results of cannabis tissue culture change much with the genotype. Before you use the dilutions, hormone doses and local regulations, make sure that they agree with the primary sources.