Element 4.24
Water quality and smart monitoring for fish
- Sheet
- 4.24
- Chapter
- 4 Food and farming
- State
- Drafted
- Release
- 16 November 2026
- Editor
- Contributed by Praecise as open research
- Reviewers
- To be named at review
- Safety
- S2, as the chapter
- Licence
- CC BY-SA 4.0
- Resources
- 11
The element
- Requirement
Dissolved oxygen, temperature, pH and ammonia are measured often enough to act before fish suffer, with readings checked against a reference instrument and kept in the farm record (4.8).
- How to tell
A measurement schedule exists for each pond, tank or cage, at least for oxygen and temperature. Readings are logged with time. Low-cost sensors are calibrated against a laboratory-grade meter at set intervals. Safe ranges for each species are taken from a species-specific source; this research found no open source giving such ranges across species, and the group should obtain them from its fish health contact or national fisheries agency. Field evidence shows why these matter: in semi-arid Tanzania, tilapia yield related to dissolved oxygen, turbidity, salinity and density (Pond water quality and fish yield in semi-arid Dodoma, Tanzania).
- Options
- Hand meters and test kits. Reliable when calibrated; labour each day.
- Low-cost sensor nodes. On Asian seabass tanks, cheap sensors with an ESP8266 Wi-Fi module reached 76 to 97 per cent accuracy against a laboratory meter over three months, but only after regression correction (Improved accuracy in IoT-based water quality monitoring for aquaculture tanks using low-cost sensors).
- Long-range radio networks. LoRaWAN sensors on buoys gave farm data online; the on-site station read about 5 °C hotter than the nearest official station 20 km away during a heat wave, showing why local measurement matters (Real-time environmental monitoring for aquaculture using a LoRaWAN-based IoT sensor network, About LoRaWAN).
- Buoys at cages. A buoy measuring oxygen, salinity, temperature and current, with server forecasting, is described with a cost of about USD 2015 (A low-cost AI buoy system for monitoring water quality at offshore aquaculture cages).
- Automatic control. Prototype pond control in Thai prawn ponds reported 93.3 per cent survival over 120 days (A prototype pond water management system for giant freshwater prawn farming in southern Thailand); predictive aerator control cut energy by 26.86 per cent in a koi pond trial (An intelligent IoT-based predictive control system for water quality and energy management in koi aquaculture).
- Computer vision. A model classified grouper feeding intensity at 99.3 per cent accuracy on its own data (Three-stream convolutional network for classifying feeding intensity of farmed fish); reviews describe vision, sonar and prediction for water quality and disease risk but give no maturity assessment (Application of artificial intelligence in aquaculture, processing, safety and traceability of aquatic products: a review).
- Settled, open, contested
Settled: oxygen and temperature need frequent measurement; cheap sensors need calibration (Improved accuracy in IoT-based water quality monitoring for aquaculture tanks using low-cost sensors). Open: how computer vision results transfer between species and farms; the feeding model's abstract shows results on one species only (Three-stream convolutional network for classifying feeding intensity of farmed fish). Contested: automation. One position holds that sensors and control save fish and energy (An intelligent IoT-based predictive control system for water quality and energy management in koi aquaculture); the other that unattended systems fail silently, so daily human checks must remain.
The four stages
Learn: what each parameter means and how sensors drift (Improved accuracy in IoT-based water quality monitoring for aquaculture tanks using low-cost sensors, Application of artificial intelligence in aquaculture, processing, safety and traceability of aquatic products: a review). Design: choose points, interval and radio link (Real-time environmental monitoring for aquaculture using a LoRaWAN-based IoT sensor network, About LoRaWAN). Build: nodes with reference calibration; wiring near water by a competent person. Deploy: log to the farm record (farmOS), check alarms, recalibrate on schedule (An intelligent IoT-based predictive control system for water quality and energy management in koi aquaculture).
Resources at each stage
1 Learn
Understand
- Real-time environmental monitoring for aquaculture using a LoRaWAN-based IoT sensor network Paper, S2
- Application of artificial intelligence in aquaculture, processing, safety and traceability of aquatic products: a review Paper, S0
- Three-stream convolutional network for classifying feeding intensity of farmed fish Paper, S0
- An intelligent IoT-based predictive control system for water quality and energy management in koi aquaculture Paper, S2
- About LoRaWAN Standard, S1
- A low-cost AI buoy system for monitoring water quality at offshore aquaculture cages Paper, S2
- Effect of stocking density of fish on water quality and growth performance of European carp and leafy vegetables in a low-tech aquaponic system Paper, S2
- Improved accuracy in IoT-based water quality monitoring for aquaculture tanks using low-cost sensors Paper, S2
- Pond water quality and fish yield in semi-arid Dodoma, Tanzania Paper, S0
- A prototype pond water management system for giant freshwater prawn farming in southern Thailand Paper, S2
2 Design
Size and choose
3 Build
Build or obtain
- Real-time environmental monitoring for aquaculture using a LoRaWAN-based IoT sensor network Paper, S2
- A low-cost AI buoy system for monitoring water quality at offshore aquaculture cages Paper, S2
- Improved accuracy in IoT-based water quality monitoring for aquaculture tanks using low-cost sensors Paper, S2
What changes between 30 adults and 300 people
At 30 adults, a calibrated hand meter and a notebook may be enough, with one sensor node as an alarm. At 300 people, several units justify a sensor network, automatic aeration and a named person who calibrates and answers alarms.
Reference designs
Open for contributions Designs with drawings, specifications and a bill of materials. Add to it
Parts list
Every resource that supports this element. Those marked External carry the label “External resource. Not made or verified by Public Protocol”.
The standing questions
Who can afford this: nodes are cheap; calibration solutions and reference meters are not. Who does the work: someone answers alarms at night. Who decides: the fish keeper sets alarm thresholds with the fish health contact.
Environments
buoys at cages (A low-cost AI buoy system for monitoring water quality at offshore aquaculture cages).
pond control in southern Thailand (A prototype pond water management system for giant freshwater prawn farming in southern Thailand).
Hot deserts and semi-arid land
water quality drives yield in ponds fed from boreholes and wells (Pond water quality and fish yield in semi-arid Dodoma, Tanzania).
Arctic cold, temperate forest, high altitude, flood plains
no source found.
Environment profiles for these settings: Small islands and coasts; Tropical rainforest and humid tropics; High humidity; Hot deserts; Savanna and drylands; Extreme heat; Arctic, subarctic and permafrost, including sea ice; Extreme cold; Temperate and boreal forests; Mountains and high altitude; River deltas and flood plains. All environments
Safety
State and evidence
Drafted
An editor has written the element's text from the resources listed under it, citing them. It is checked when it is reviewed.
Evidence from a real place or system is recorded here, with what did not work. What the states mean
History
- Drafted in the release of 16 November 2026.