UA50 Spray-Width Acceptance: How to Verify Pass Spacing Before Signing Off a Purchase

By UNITED UAV Official

Commercial disclosure: UNITED UAV sells the UA50. This article proposes a procurement measurement framework. It does not present a UA50 flight trial, a pesticide application prescription, or an independently validated operating envelope.

Original seller product image of the UA50. The image identifies the aircraft; it is not evidence of a completed spray-pattern test.

Turn a width label into a testable purchasing question

The current UA50 comparison panel lists a 50 L pesticide tank and an “Effective Spray Width” of 8–10 m. Those are seller declarations about the offered equipment. They do not identify the crop target, sampling method, aircraft configuration, weather window, or acceptance statistic behind a particular pass spacing. A buyer can use the width range as a starting reference while still requiring evidence that the proposed configuration produces an acceptable overlapped distribution under the intended evaluation conditions.

The useful purchasing question is therefore more specific than how wide the aircraft can spray. It is whether a proposed distance between adjacent flight lines can be defended with repeatable measurements, and whether the supplier can explain the conditions and uncertainty attached to that answer. A visible wet patch at the edge of a pass, a controller width setting, and a photograph of a demonstration each answer different questions. None should silently become a statement that the complete area between neighbouring routes received a uniform application.

The framework below is intended for a buyer and a qualified supplier test team agreeing on a demonstration before accepting an aircraft configuration. Initial evaluation should use a suitable clean-water procedure approved for the equipment and the test site. Actual application choices remain tied to the equipment instructions, the permitted operating conditions and the intended application. The proposed calculations explain how to assess a report; they do not prescribe a flight height, chemical concentration or operational spray setting.

Freeze the configuration before collecting a convincing photograph

A swath report belongs to a particular system configuration. Record the aircraft identifier, spray assembly revision, atomizer or nozzle identity, mounting arrangement, controller software, relevant rate-control mode and the source of the proposed spacing. A statement such as “UA50, standard setup” leaves too much room for substitutions. The same model name can appear on a quotation, a product photograph and a test log while the parts or control settings relevant to distribution differ.

Keep the application configuration separate from optional equipment. The UA50 listing offers an optional solid spreader and an optional ground generator configuration. Their presence in a catalogue does not establish that they were installed for a liquid-distribution trial, or that a generator powers propulsion in flight. A liquid-spray acceptance record should identify the liquid system actually tested. A spreading demonstration requires its own material, collection and interpretation method; its result cannot fill a gap in a liquid-spray report.

Loading also deserves a precise description. Record the measured or otherwise documented starting liquid quantity, the ending quantity and the corresponding aircraft loading condition. Use loading cases permitted by the supplier for that exact configuration. Comparing a demonstration near the beginning of a tank with a later demonstration after considerable liquid has been dispensed can confound the question being investigated. A 50 L tank label does not establish the permitted loading condition for every liquid density, attachment or test arrangement.

Define what each measurement can actually establish

Purdue Pesticide Programs, in PPP-154 chapter 6, distinguishes effective pass spacing from the outer reach of a spray pattern. Its hydraulic-nozzle sizing discussion also expressly excludes rotary atomizer types. The relevant lesson here is to require an appropriate measurement basis, rather than transferring a nozzle recipe or a generic percentage reduction into a UA50 acceptance claim. An operator should obtain the procedure applicable to the supplied spray assembly.

Separate distribution, recovered quantity and biological outcome in the report. A collection system can show how its measured response varies across a transect. A tank or flow measurement can indicate how much liquid left a defined system boundary. Neither observation alone establishes target retention, penetration into a canopy, droplet spectrum or pest control. When the procurement objective requires one of those additional outcomes, specify a suitable additional measurement instead of allowing a photograph or a controller total to stand in for it.

Choose one primary response before testing. It could be a suitably validated mass, volume or tracer-recovery measurement from equal collection areas. Water-sensitive papers can provide useful pattern observations, but image stain coverage should not automatically be treated as recovered liquid mass. Likewise, an image-analysis output is only as interpretable as its sampling, handling, thresholding and calibration. Changing response units partway through a comparison makes a report harder to audit, even when every chart looks technically polished.

Put the sampling boundary in writing

The test plan should define the target plane, the measured cross-track coordinates, the collection area associated with each position and the range over which the analysis will be performed. Height should identify its reference surface; a value measured above bare ground is not automatically a value above a crop target. Coordinates should have a stated relationship to the recorded aircraft track. These definitions let a reviewer distinguish a displaced pattern from a narrow pattern, and a navigation offset from an apparent sampling offset.

The sampling span should be chosen to observe the intended pattern and its edges using a procedure suitable for the test. Do not choose an analysis interval after looking at the results merely because a narrower interval removes inconvenient low readings. If the evaluation includes an interior treatment region and separate boundary behaviour, define both regions in advance. Retain the excluded observations with their coordinates and reasons for exclusion so that another analyst can reconstruct the result.

Include method controls that suit the chosen response. For a tracer method, these may include blanks, handling controls and a documented way to relate instrument response to the recovered quantity. For a collection method, record balance resolution, collection area and the handling of evaporation or other relevant loss. A supplier should be able to explain the uncertainty that matters at the size of the observed differences. Extra decimal places in a spreadsheet do not compensate for an uncertain collection or measurement process.

Design repetitions to expose variation, not hide it

One clean-looking pass is weak evidence for a repeatable operating result. In their USDA Agricultural Research Service publication on measurement and analysis methods for effective swath width, Fritz and Martin identify substantial variation among repeated unmanned spray passes and describe the value of additional repetitions and analysis that accounts for that variation. Their work supports repeated measurement; it does not validate a UA50 configuration or supply a universal acceptance threshold for this purchase.

Agree on a bounded trial matrix before the demonstration. It should identify the proposed spacing, the configuration held constant, the relevant loading cases and the environmental conditions to be recorded. Where the permitted test plan includes different headings or cases, preserve their identities. A report that averages all cases into one attractive curve can conceal a repeatable directional offset or a loading-sensitive change. A purchaser needs to know which cases support the proposed setting and which do not.

Record for each test run	Why the buyer needs it
Configuration and loading case	Establishes which supplied arrangement the result describes
Planned and actual track, time and direction	Separates geometry, execution and timing errors
Relevant environmental observations	Preserves the conditions under which the result was obtained
Raw sample position and response	Allows the transverse pattern to be reconstructed
Instrument and processing revision	Makes the reported calculation reproducible
Deviations and unusable observations	Prevents silent selection of favourable results

Pair the environmental record with the time of each run. A single wind observation taken at the start of a long demonstration may not describe every later pass. Specify how the test team will identify a run affected by conditions outside its agreed window. Preserve such a run as a documented exception instead of quietly reclassifying it as a successful replication. The same principle applies to interrupted collection, an instrument problem or a missing track log.

Randomising or otherwise controlling the order of permitted cases can help avoid confusing gradual changes with a configuration effect. For example, if every high-loading run is performed first and every lower-loading run occurs much later, the report cannot easily separate loading from the change in time and conditions. The exact experimental design belongs to the qualified test team, but the procurement record should state how this confounding was considered rather than leaving the buyer to infer it from timestamps.

Evaluate the overlapped pattern, not only a single pass

For a proposed route spacing, the relevant interior distribution is the combination of neighbouring passes. A single-pass plot is useful input, but the report should explain how adjacent patterns were combined and which assumptions were made about alignment and repeatability. A calculation that shifts copies of one measured pass assumes that subsequent passes behave like that selected example. That can be a useful exploratory model, provided it is identified as a model and compared with repeated observations.

The buyer should ask to see both the individual runs and the resulting combined distributions. Where the proposed analysis pools repeated passes, it should retain enough run-level information to expose the range of plausible outcomes. Do not collapse everything into an average before checking offsets or asymmetry. A repeatable lateral displacement can disappear into a smooth pooled curve even though it still matters at a boundary or when adjoining routes are flown in a particular sequence.

Specify which positions enter the summary statistic and how unequal sample spacing or collection area is handled. Uniformly spaced, equal-area observations are simpler to interpret than a mixture of intervals and collection sizes. If weighting or interpolation is needed, document it. More importantly, preserve the raw observations alongside the processed grid. A buyer should be able to distinguish an observed low-response strip from a low value created by an interpolation choice.

Use statistics that reveal what an average can conceal

A mean describes the overall response across the selected evaluation region. It says little about where low or high values occur. A useful report pairs the mean with the transverse plot, the observed range, relevant low-response positions and a documented measure of variability. The coefficient of variation, commonly expressed as a percentage, is one possible descriptive statistic: divide the chosen standard deviation by the mean, then multiply by 100. State the standard-deviation convention and the population being described.

Consider two deliberately invented sets of six equal-area observations. The values below are relative units for explaining a calculation. They are not UA50 test results, collected spray data or proposed application rates. Both rows sum to 60 and have a mean of 10. Using the sample standard deviation, with the squared-deviation sum divided by five, produces very different variability estimates.

Illustrative response	Position 1	Position 2	Position 3	Position 4	Position 5	Position 6	Sample CV
Example A	9	10	10	10	10	11	6.32%
Example B	6	8	10	10	12	14	28.28%

The equal means would conceal the difference if the report contained only a total or average. In Example B, the low readings sit on one side and the high readings on the other; that spatial structure matters. A single variability percentage still does not show whether a poor result is a broad gradient, a narrow strip, an isolated sampling error or an edge effect. This is why the plotted positions and the original run identity must remain attached to the statistic.

Do not turn either illustrative percentage into a universal acceptance limit. The appropriate criterion depends on the measurement method, target, intended evaluation and agreement between the parties. Near-zero means also make a relative variability statistic unstable or uninformative. A missing observation is not automatically a zero, and a genuine zero is not automatically an outlier to delete. Both classifications require a reason that can be checked against the field record.

Keep spacing, flow and application volume consistent

A spacing decision also changes the area represented by a flight line. For a dimensional illustration, let total liquid flow be Q litres per minute, ground speed be v metres per second, spacing be W metres and nominal application volume be R litres per hectare. For uninterrupted straight travel with the stated geometric assumptions, the relationship is Q = 0.006 × R × v × W. The constant follows from converting minutes to seconds and square metres to hectares; it is not a UA50 performance factor.

Using hypothetical values of R = 20 L/ha, v = 4 m/s and W = 8 m gives Q = 3.84 L/min. If flow and speed stayed unchanged while the represented spacing became 10 m, the nominal geometric application volume would become 16 L/ha. These numbers are a calculation example, not suggested settings for the aircraft or a crop. Their purpose is to show why a spacing change cannot be assessed in isolation from flow and speed. They say nothing about actual deposited quantity or the equipment's ability to deliver the assumed flow.

If the supplied controller changes output when the spacing or target rate changes, document that behaviour from the applicable instructions and the actual test record. Do not assume which compensation functions exist or whether they were active. The requested, displayed and independently observed quantities should remain distinguishable. A flow-meter total can be consistent with its control logic while the spatial distribution still needs attention; conversely, a visually even pattern does not establish that the required total quantity was delivered.

Agree on the decision rules before reviewing the result

A procurement acceptance clause should state the configuration, the applicable evaluation conditions, the response being measured and the decision rule. It should also define what happens when the result is close to the limit or the uncertainty is too large to support a clear decision. A simple passed flag without the criterion, raw evidence and identified conditions gives the buyer little protection against misunderstanding and gives the supplier little clarity about what demonstration is expected.

Allow for a conditional result. A report may support a narrower set of evaluation conditions than the buyer initially requested, require a repeat of an invalid run, or identify an unresolved measurement uncertainty. These outcomes are more informative than forcing every demonstration into an unconditional pass or fail. If a configuration is changed after a disappointing result, give it a new revision identity and identify which comparisons remain valid. Reusing the old label would obscure what was actually corrected.

The final deliverable should let another technically competent reviewer reproduce the conclusions without attending the demonstration. Request the raw data in a usable format, coordinates, run identifiers, processing equations, software or spreadsheet revision and an exception log. Retain the supplier's signed or otherwise attributable explanation of how the proposed spacing was derived. Photographs and video can support this record by showing the arrangement, but they should accompany the quantitative evidence rather than replace it.

Ask for the evidence that applies to the aircraft being quoted

For the UA50 Agricultural Drone - 50L Model for Precision Farming, a useful enquiry asks for the exact liquid-system configuration, the basis of the proposed working spacing and the test conditions under which the supplier would demonstrate it. Include the target measurement objective and the intended receiving surface or crop context. This lets the supplier identify the relevant evidence or explain what still needs to be evaluated.

The purchasing decision then has a clearer basis. The catalogue identifies the offered aircraft and its declared reference values. The trial record identifies what a particular configuration demonstrated, with measurable limitations. The acceptance agreement identifies which of those observations satisfy the buyer's stated evaluation. Keeping those three records connected makes a width claim easier to challenge, reproduce and use responsibly, without turning a product label into a field result that nobody actually measured.