Sea Drones for Port Water-Quality Monitoring: What to Check Before Commissioning a Survey

Sea drones can extend a port’s monitoring coverage, but useful evidence depends on the survey design, sensors and interpretation. Before commissioning a water-quality survey, establish the question, operating conditions and deliverables that will help your team make a defensible decision.

A harbour manager may want to understand a recurring patch of cloudy water, compare conditions around an outfall or investigate low oxygen in a sheltered basin. These are different questions. They should not automatically receive the same survey package simply because a supplier has a capable vessel available.

The purchasing decision starts with evidence, then equipment. Write down what the findings could change: an inspection, further sampling, a maintenance programme or a conversation with a tenant. This keeps the project focused when a demonstration offers more measurements than the port actually needs.

Define the water-quality problem first

Describe the suspected issue, its location and the circumstances in which it appears. Include rainfall, tide, dredging activity or vessel movements where relevant. Separate observations from assumptions. Cloudy water is an observation; attributing it to a particular business requires additional evidence.

Choose a primary question narrow enough to answer. For example: does turbidity vary across this basin at comparable tidal stages? Or: where should laboratory samples be collected after rainfall? Agree what result would trigger follow-up and who will decide whether the evidence is adequate.

A drone survey can contribute to an environmental investigation without resolving every part of it. If the concern is a named chemical, specify an appropriate analytical approach. General water-quality sensors should not be sold as a substitute for identifying substances they do not measure.

Which type of sea drone suits the task?

Sea drone is a broad term. Cleaner Seas’ guide to sea drones and their uses introduces the main categories. A commissioning brief should identify the actual platform and explain why it fits the port’s question.

Uncrewed surface vessels

A surface vessel can carry suitable sensors along planned routes, potentially helping map conditions across accessible water. Ask where the sensors sit and what depth they represent. Readings near the surface do not automatically describe conditions near the seabed.

Autonomous underwater vehicles

An AUV operates below the surface and may carry instruments for the water column or seabed. The National Oceanography Centre’s AUV overview illustrates research capabilities. Those examples do not establish suitability for a congested harbour: assess the proposed vehicle, navigation and recovery arrangements specifically.

Tethered remotely operated vehicles

An ROV can support targeted underwater observation or measurement where its payload and access allow. Its tether, launch position and support team constrain the working area. Ask how the operator will manage structures, currents and nearby vessel activity.

Aerial drones

Aerial imagery can document visible surface features and guide further investigation. It does not directly measure conditions throughout the water column. Specialist sampling attachments need their own validation, handling procedures and operating permissions rather than assumptions based on the camera’s performance.

Compare drones with other monitoring methods

A fixed sensor can record changes at one location over time. A mobile survey can cover more locations during a deployment. Crewed sampling may offer practical flexibility and direct sample handling. Often, the strongest design combines methods instead of choosing one technology for every task.

Compare staffing and repeatability as carefully as spatial coverage. A route is useful only if the team can collect comparable evidence next time. Ask each provider to price the same question and deliverables, making differences in scope visible before comparing costs.

Specify measurements and their limitations

List the parameters required and why each matters. Temperature, conductivity, dissolved oxygen, pH and turbidity can describe different aspects of conditions. Ask for units, sensor ranges and relevant uncertainty. Avoid a large package of measurements with no agreed interpretation or management purpose.

Turbidity describes an optical property associated with water clarity; it is not a chemical identification. A high reading does not independently name the source or establish toxicity. Similarly, a fluorescence signal needs appropriate interpretation before it becomes a claim about a particular biological risk.

The USGS guidance on multiparameter instruments explains measurement and instrument considerations. Use competent survey advice to translate those principles into the proposed marine deployment. Require the provider to explain any indirect estimates rather than presenting them as direct laboratory measurements.

Require calibration and field quality checks

Ask for sensor identities, calibration records and the proposed checks before and after deployment. Establish how the surveyor will recognise fouling, drift, damage or implausible readings. A precise-looking decimal on a dashboard is not evidence that a measurement is reliable.

USGS monitoring guidance emphasises cleaning, calibration, field observation and record review. Although written for continuous monitoring, it reinforces a useful purchasing principle: the quality process belongs in the specification, alongside the sensors themselves.

Specify how checks will appear in the report. Request flags for suspect data and an explanation of exclusions or corrections. The port should be able to distinguish a measured environmental change from a sensor problem without reconstructing the supplier’s entire processing workflow.

Design coverage around tides, depth and timing

Agree survey boundaries, routes, measurement depths and repeat visits with the survey team. Include enough context to interpret spatial differences. If one area is measured early and another several hours later, changing conditions may complicate comparisons across the resulting map.

Record tide, weather and relevant port activity during collection. Ask how currents could move the feature under investigation. For a rainfall question, define what counts as the sampling event and whether a comparison under different conditions is needed.

Make missed coverage visible. A cancelled route or inaccessible berth should remain a gap in the report, not disappear inside an apparently complete map. Agree how the provider will notify the port and whether returning to complete the work is included.

Protect sample handling and independent confirmation

If the drone collects water samples, specify collection equipment, containers, depth, labelling, preservation and transport with the laboratory. Record collection times and locations. Different analyses require different handling, so a generic bottle collection procedure is unlikely to answer every monitoring question.

Ask whether duplicates, blanks or other quality controls are appropriate for the study. Arrange selected comparison measurements or laboratory analysis where needed to check the interpretation. This confirmation work should be priced openly rather than appearing later as an unexpected extra.

Agree the chain of custody and who owns the samples and results. If findings may inform a regulatory submission, establish the relevant acceptance requirements before deployment. A commercially useful survey and an accepted compliance dataset are not necessarily the same deliverable.

Check permissions and practical operating limits

Discuss the operation with the harbour team and identify the permissions required for the actual platform and location. Define navigation responsibilities, communications, supervision, launch and recovery, and interaction with other users. Do not assume a small uncrewed vessel can operate unattended in busy waters.

Require a plan for lost communications, equipment failure and recovery. Ask who can suspend the survey and how vessel traffic affects the route. Weather and current limits should describe the proposed operation, with clear arrangements for postponement and remobilisation.

Plymouth Marine Laboratory’s autonomy research places these technologies within wider observation and testing work. Autonomy describes how a system operates; AI describes particular computational methods. Neither term independently demonstrates safe deployment or reliable environmental interpretation.

Agree deliverables and the full project cost

Request raw readings, processed data, locations, times, units, quality flags and an explained report in usable formats. Establish ownership, access rights and any continuing software fees. The port should retain enough information to compare surveys even if it later changes supplier.

Price planning, mobilisation, staffing, permissions, calibration, laboratory work, interpretation and repeat visits. Compare the completed project cost rather than the daily vehicle rate. Claims of savings should reflect a comparable local scope, not an unrelated demonstration in different conditions.

Acceptance criteria should describe delivered coverage and evidence quality, not merely hours afloat. Name the person who reviews the report and records follow-up decisions. This connects survey spending with the inspection, maintenance or further investigation it was commissioned to support.

Test the brief before extending the programme

Consider a hypothetical marina investigating cloudy water near one drainage outlet. Start with a defined comparison between locations and conditions, supported by appropriate samples. Ask the surveyor to explain which measurements could distinguish the working hypotheses and which would simply describe the visible pattern.

After the initial deployment, review coverage gaps, sample results and uncertainty with the environmental manager. If the evidence points towards a land-based source, a drainage inspection may be the next useful action. More drone routes would not automatically provide the missing answer.

Only extend the programme when the first report demonstrates how repeat surveys will inform a decision. Document any changes to routes, instruments or timing so later comparisons remain intelligible. A modest survey with a clear consequence can be more useful than an extensive dataset without an agreed owner.

Frequently asked questions

Can sea drones identify pollution automatically?

Only within the capabilities and limitations of their instruments and analysis. General sensor readings can guide investigation but do not automatically identify a pollutant, its source or its ecological effect.

Are drone surveys cheaper than crewed surveys?

Sometimes, depending on scope, access, staffing and repeat work. Compare full quotations for the same evidence requirement. No universal saving applies to every port or monitoring question.

Can one survey prove water quality has improved?

One survey documents particular conditions. Demonstrating improvement requires a suitable comparison design and consideration of other influences. Agree that design before commissioning a project intended to measure change.

Should a port buy a drone or commission a service?

Assess workload, skills, maintenance and data interpretation first. A service can test whether the method answers the question before the port commits to owning and supporting equipment.